Datasets:

shreyaspimpalgaonkar

/

ycombinator_s24

like 0

Elevate

elevate-2

https://www.useelevate.dev/

New York, NY, USA; Remote

We make it easy for AI agents to buy anything on the internet. For example, if a user asked an AI agent to buy a book on an e-commerce site, we handle the purchase via our Purchasing API. AI agents are poised to elevate the way we shop, but they aren’t reliably making purchases today. We’re changing that by bringing uniformity to non-uniform checkout flows across the internet. Our expertise is at the intersection of payments and e-commerce. Angelica was a staff engineer at Stripe, driving API evolution, design, and governance across payments products. Scott was a senior investor at Sycamore Partners, investing in and partnering with leading consumer companies. Our solution will enable AI agents to make payments of every kind, including e-commerce (e.g., Amazon, eBay), consumer services (e.g., Uber, DoorDash), and B2B (e.g., supplier payment). In the US e-commerce market alone, there are 10B transactions every year, representing a significant opportunity for us. We have already onboarded multiple AI agent customers who have thousands of monthly users and have begun integrating them to power their e-commerce purchases.

Financial infrastructure for AI agents

Fintech

Fintech -> Payments

S24

Active

Early

https://www.ycombinator.com/companies/elevate-2

https://yc-oss.github.io/api/batches/s24/elevate-2.json

Title: Elevate: Financial infrastructure for AI agents | Y Combinator URL Source: Markdown Content: ### Financial infrastructure for AI agents We make it easy for AI agents to buy anything on the internet. For example, if a user asked an AI agent to buy a book on an e-commerce site, we handle the purchase via our Purchasing API. AI agents are poised to elevate the way we shop, but they aren’t reliably making purchases today. We’re changing that by bringing uniformity to non-uniform checkout flows across the internet. Our expertise is at the intersection of payments and e-commerce. Angelica was a staff engineer at Stripe, driving API evolution, design, and governance across payments products. Scott was a senior investor at Sycamore Partners, investing in and partnering with leading consumer companies. Our solution will enable AI agents to make payments of every kind, including e-commerce (e.g., Amazon, eBay), consumer services (e.g., Uber, DoorDash), and B2B (e.g., supplier payment). In the US e-commerce market alone, there are 10B transactions every year, representing a significant opportunity for us. We have already onboarded multiple AI agent customers who have thousands of monthly users and have begun integrating them to power their e-commerce purchases. Elevate Founded:2024 Team Size:2 Location:New York ### Active Founders ### Scott Cara, Founder Scott is the co-founder and CEO of Elevate AI. Scott was previously a consumer investor at Sycamore Partners where he partnered with c-suite executives at leading companies to drive growth. Prior to Sycamore, Scott was a management consultant at Bain & Co. Scott studied Computer Science and Mathematics at Duke University and holds an MBA from Harvard Business School. ### Angelica Schwartz, Founder Angelica is the co-founder and CTO of Elevate. Prior to Elevate, Angelica was a Staff Software Engineer at Stripe on the Standard Library team, which owned API design and governance of core payments APIs across the company. Before that, she was a Technical Lead at Google on Ads, launching new products on massive scale properties such as Search and YouTube. Angelica studied Computer Science and Mathematics (minor) at Duke University, and holds a Master of Engineering Management also from Duke.

Vendra

vendra

https://vendra.io/

San Francisco, CA, USA

Vendra uses AI to automate project management and information lookup when working with vendors and suppliers - all with no change to existing workflows - letting hardware teams focus on building and shipping amazing products

AI copilot for hardware teams working with vendors

B2B

B2B -> Engineering, Product and Design

S24

Active

Early

https://www.ycombinator.com/companies/vendra

https://yc-oss.github.io/api/batches/s24/vendra.json

Title: Vendra: AI copilot for hardware teams working with vendors | Y Combinator URL Source: Markdown Content: ### AI copilot for hardware teams working with vendors Vendra uses AI to automate project management and information lookup when working with vendors and suppliers - all with no change to existing workflows - letting hardware teams focus on building and shipping amazing products Vendra Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Shan Mohta, Founder CEO of Vendra. I built cool hardware at Skydio, Apple, Microsoft. ### Anish Bhardwaj, Founder CTO of Vendra. I build software. ### Company Launches [### Vendra - AI vendor and project management for hardware teams]( ### TL;DR * Hardware engineers **_should_** spend their time building cool products, but they **_actually_** spend their time dealing with vendors and doing non-engineering work (digging through emails, searching for files, updating spreadsheets, etc.) * Traditional project management solutions don’t really work in hardware, and vendor communications inevitably revert to emails/text messages/WeChat/etc. * Vendra uses AI to understand these vendor communications, giving hardware teams modern project management and instant information lookup with almost no change to their existing workflow * We offer solutions to hardware teams at all stages - [book a demo today!]( ### The Team We’re [Shan]( and [Anish]( We’ve known each other since childhood when our parents moved from India to the same apartment complex in Santa Clara. Between us, we’ve built a lot of cool hardware (cameras for autonomous drones, AR/MR headsets) and cool software (ML for satellites, AI for the shipping industry). ### The Problem As a former hardware engineer, Shan experienced the challenges of managing complex projects across a network of vendors, suppliers, and manufacturers. He saw hardware teams struggle to use traditional project management solutions with vendors because of the unstructured nature of projects and communications, security and access issues, and difficulties in the adoption of new processes. The workflow inevitably defaulted to emails, text messages, WeChat, etc., often leading to misplacement of crucial information. A 2021 report revealed that 70% of engineers still use email for external collaboration, struggling to track critical information across teams. 94% of engineers want faster access to data, and 29% often work with outdated information, leading to costly delays, reworks, and longer time to market. We’re building Vendra to change that. ### The Solution **Vendra uses AI to bring a smart project management suite and file repository to hardware teams working with vendors and suppliers - all with no change to existing workflows.** Simply connect Vendra to your email/WeChat/WhatsApp/Slack (or cc it to your conversations and threads): Vendra automatically contextualizes your emails/messages/files and generates key takeaways like project and conversation summaries, open items, blockers, tasks, etc. It also creates a smart file directory so you never have to go digging through emails or attachments ever again. Easily access all this information from anywhere through our SlackBot, GPT-like chat interface, or web-based dashboard: You can ask Vendra AI things like * ‘Show me all the quotes and lead times we got from vendors for the upper bracket machining in the past month’ * ‘What are all the open items for the enclosure injection molding tooling?’ * ‘Pull up the RF module datasheet the supplier sent me last week’ * And so much more! For teams looking for 2-way external collaboration, we also offer VendraPM - a secure end-to-end platform to onboard your vendors and work together across projects and tasks. Features like secure file sharing, document revision control, and AI-enabled language translation make external work easier than ever: ### The Ask * If you work in hardware, [book a demo today!]( * We also love warm intros to folks in hardware - if you know anyone working in (or around) hardware, we’d love to chat with them to understand their workflow and see how we can help! * Follow us on [LinkedIn]( and [X!]( #### YC Sign Photo

Guardian RF

guardian-rf

https://guardianrf.com/

San Francisco, CA, USA

Cost-Effective Drone Detection and Signal Intelligence for 21st- Century Peacekeeping.

Drone detection and signal intelligence for US defense.

Industrials

Industrials -> Drones

S24

Active

Early

https://www.ycombinator.com/companies/guardian-rf

https://yc-oss.github.io/api/batches/s24/guardian-rf.json

Title: Guardian RF: Drone detection and signal intelligence for US defense. | Y Combinator URL Source: Markdown Content: ### Drone detection and signal intelligence for US defense. Cost-Effective Drone Detection and Signal Intelligence for 21st- Century Peacekeeping. Guardian RF Founded:2024 Team Size:3 Location:San Francisco ### Active Founders ### Lucas Raskin, Founder Physics enthusiast. Building next generation counter UAS capabilities. ### John Andrzejewski, Founder Former Physics at Georgetown. Working on counter UAS and hitting PRs on bench press. ### Eli Kerstein, Founder My background is in physics and comp sci. I build Counter UAS systems.

Biocartesian

biocartesian

https://www.biocartesian.com

San Francisco, CA, USA

Finding new cures requires seeing what and where the abnormal molecules are in a diseased tissue, but current tools see less than 1% of those molecules. Biocartesian combines microscopy and new chemistries to see 50X more, offering unprecedented insights into disease biology and new therapies.

Converting biology into digital molecular maps

Healthcare

Healthcare -> Drug Discovery and Delivery

S24

Active

Early

https://www.ycombinator.com/companies/biocartesian

https://yc-oss.github.io/api/batches/s24/biocartesian.json

Title: Biocartesian: Converting biology into digital molecular maps | Y Combinator URL Source: Markdown Content: ### Converting biology into digital molecular maps Finding new cures requires seeing what and where the abnormal molecules are in a diseased tissue, but current tools see less than 1% of those molecules. Biocartesian combines microscopy and new chemistries to see 50X more, offering unprecedented insights into disease biology and new therapies. Biocartesian Founded:2023 Team Size:2 Location:San Francisco ### Active Founders ### Ian Dardani, Founder Co-founder of Biocartesian. Microscopy + molecular tools + engineering. Previously Bioscience Enterprise at Cambridge, Bioengineering PhD at UPenn. ### Jess Li, Founder Co-founder of Biocartesian. Genetics + epigenetics, next-gen sequencing. Previously PhD at UPenn, BD at imvaria. ### Company Launches [### Biocartesian - Converting biology into digital molecular maps]( **Tl;dr**: Finding new cures requires knowing **what** and **where** the abnormal molecules are in a diseased tissue, but current tools see less than 1% of those molecules. [Biocartesian]( combines microscopy and new chemistries to see a comprehensive picture, offering unprecedented insights into disease biology and new therapeutic approaches. **👩‍🔬 Team** -------------- We’re [**Ian**]( and [**Jess**]( two former UPenn PhDs on a mission to see every molecule in biological specimens. We’ve spent years advancing and troubleshooting the latest tools in biology, from DNA sequencing to microscopy, but even with these efforts, these tools revealed only a fraction of the underlying biology. We dreamed of a way for scientists to see everything (and fast) so they could reduce their guesswork and focus on the most promising therapeutic approaches. ❌ **Problem** ------------- Most diseases remain a mystery because existing methods can see less than 1% of the molecules within a biological specimen. These methods can also take months for scientists to set up and often fail, wasting thousands of dollars per specimen and millions in failed drug programs. 🔬 **Our approach** ------------------- We’re combining new chemistries and microscopy to see a full map of the molecules in biological specimens. Imagine if we could see into the ~500 million existing tumor specimens to decode these cancers. To use our technology, scientists simply ship their biological specimens to our laboratory, where we automate their conversion into digital molecular maps – saving them months of tedious data collection and troubleshooting. Our solution: * Enables scientists in any disease area to 50X their data collection * Saves scientists months of time * Lowers the cost of biological measurements * Expands access to bespoke assays developed by any scientist by integrating with our platform * Meets the growing demand for biological data as AI-led analysis supersedes the bottleneck of human-led analysis * Turns biology from 99% invisible to an interpretable problem **🙏 Asks** ----------- If you’re excited about this and would like to learn more, email us at: [**[email protected]**](mailto:[email protected]) or reach us at [

Honeycomb

honeycomb

https://honeycomb.sh/

San Francisco, CA, USA

Honeycomb is building agents to automate tasks across the entire software development lifecycle. Our agents are state-of-the-art, achieving 22% on the SWE-Bench coding benchmark.

Autonomous Software Agents

B2B

B2B -> Engineering, Product and Design

S24

Active

Early

https://www.ycombinator.com/companies/honeycomb

https://yc-oss.github.io/api/batches/s24/honeycomb.json

Title: Honeycomb: Autonomous Software Agents | Y Combinator URL Source: Markdown Content: Honeycomb: Autonomous Software Agents | Y Combinator =============== []( "Y Combinator") [About]( [What Happens at YC?]( Interview Guide]( Blog]( [Companies]( [Startup Directory]( Companies]( Directory]( YC]( [Startup Jobs]( [All Jobs]( Engineering]( Operations]( Marketing]( Sales]( Internship Program 2024]( Job Guide]( Coaching]( Startup Jobs Blog]( [Find a Co-Founder]( [Library]( [SAFE]( [Resources]( [Startup School]( Investors]( News]( Open main menu Apply for **W2025** batch.[Apply]( "Apply for W2025 batch.") [Home]( Honeycomb ========= Autonomous Software Agents [S24]( Active [b2b]( Francisco]( * * * [Company]( [Jobs](  [ * * * ### Autonomous Software Agents Honeycomb is building agents to automate tasks across the entire software development lifecycle. Our agents are state-of-the-art, achieving 22% on the SWE-Bench coding benchmark. Honeycomb Founded:2024 Team Size:2 Location:San Francisco Group Partner:[Jared Friedman](  []( "LinkedIn profile") []( "Twitter account") []( "Github profile") ### Active Founders ### Andrew Liu, Founder Hi, I'm Andrew! CS & EECS + Math at MIT. Andrew Liu [Honeycomb](  []( "LinkedIn profile") ### Ishank Agrawal, Founder I'm Ishank, CS @ MIT. Ishank Agrawal [Honeycomb](  []( "Twitter account") []( "LinkedIn profile") ### Company Launches [### 🍯 Honeycomb — Bringing autonomy to software engineering]( **TLDR: Honeycomb is building agents to help automate tasks across the entire software development lifecycle.** We’re building agents that can automate tasks across the software development lifecycle—bug fixing, pull request reviews, knowledge retrieval, and more. Our coding agents are state-of-the-art on SWE-bench, a benchmark for evaluating AI-driven workflows on GitHub issues. **The Problem** On average, developers spend more than a third of their time on maintenance, testing, documentation, code review, and other tedious work. This work is necessary but contributes to slower cycles—this problem is exacerbated with larger engineering teams. While many large companies are working to automate their development processes with AI, this integration is challenging. Creating robust, functional workflows requires significant time and effort. We're developing a solution that works for everyone. **The Solution** We’ve built a coding engine that allows our agents to make efficient, precise, and accurate changes to large codebases. [ Here, we’ve asked our agent to make some quick improvements to its own UI. This is just one use case for our engine; we’re capable of implementing agents across the entire software stack, from testing, documentation, and code reviews to more complex bug fixes and even 0-to-1 development. We’ve been working hard to ensure that our agents deliver on real-world issues. We evaluated our system on SWE-Bench, a dataset that tests the capabilities of agentic systems against GitHub issues: Our engine resolved **22.06% of the 2,294** issues in SWE-bench's test set, surpassing the previous state-of-the-art Amazon Q Developer Agent (19.75%). We additionally resolved **40.6% of the 500** issues in SWE-bench’s verified set, also surpassing previous state-of-the-art (38.8%). **Our Initial Set of Developer Tools Includes:** * Automated bug fixing from issue tickets * Automated bug tracing from code reviews * Automated code reviews on pull requests * Knowledge search and sharing Honeycomb also seamlessly integrates with GitHub, GitLab, Slack, and Discord— with more to come—allowing teams to solve bugs or query their code with just a single command. **The Team** 👋 Hey, it’s [Andrew]( and [Ishank]( We met during our first year at MIT. Since then, we’ve done research spanning statistics, machine learning, algorithms, and astrophysics. Through research and side projects, we’ve spent a lot of time thinking about ways to program better and eventually realized that the intersection between AI and programming is a future that we’d like to build towards. **Our Ask** If you’ve encountered repetitive or tedious problems in software development, we’d love to talk! We would also love to talk to any engineering managers looking to accelerate their development process. Reach us at [[email protected]](mailto:[email protected]), or join the [waitlist]( Footer ------ [Y Combinator]( "Y Combinator") ### Programs * [YC Program]( * [Startup School]( * [Work at a Startup]( * [Co-Founder Matching]( ### Company * [YC Blog]( * [Contact]( * [Press]( * [People]( * [Careers]( * [Privacy Policy]( * [Notice at Collection]( * [Security]( * [Terms of Use]( ### Resources * [Startup Directory]( * [Startup Library]( * [Investors]( * [SAFE]( * [Hacker News]( * [Launch YC]( * [YC Deals]( ### Make something people want. [Apply]( [Twitter]( © 2024 Y Combinator

Saldor

saldor

https://saldor.com/

San Francisco, CA, USA

Engineers use Saldor's API to provide high quality web data to their AI agents.

The web scraper for AI

B2B

B2B -> Engineering, Product and Design

S24

Active

Early

https://www.ycombinator.com/companies/saldor

https://yc-oss.github.io/api/batches/s24/saldor.json

Title: Saldor: The web scraper for AI | Y Combinator URL Source: Markdown Content: ### The web scraper for AI Engineers use Saldor's API to provide high quality web data to their AI agents. Saldor Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Jacob Bland, Cofounder, CEO Cofounder and CEO at Saldor. Priorly product manager and software engineer for Crusoe, the AI cloud. Before that, on the data team at Square. ### Jack Cameron, Founder/CTO Cofounder and CTO at Saldor. Previously at Crusoe Energy building storage solutions for Crusoe Cloud. ### Company Launches [### Saldor - The web scraper for AI]( Hey YC! We’re Jack Cameron and Jacob Bland at Saldor. **TL;DR:** **Saldor is the web scraper for AI.** ### **The Problem** LLMs need quality and recent internet data to solve specific problems—that’s why RAG is so popular! But gathering this data from the internet is hard. Orchestrating crawlers, finding the pages of interest within a site, maintaining the context from page layouts, etc., can be challenging. Updating the store as this data changes over time can be costly and time-consuming as well. ### **The Solution - Saldor!** Saldor intelligently crawls sites and extracts the content. With just a few lines of code, engineers can convert messy web data into clean, ready-to-use output, whether it's human-readable text for LLMs or structured JSON for traditional programs. ### **Who are we?** Two engineers who spent years building AI infra at Crusoe Cloud. We worked with teams like Together  AI and Mosaic ML to accelerate the training and deployment of new models. These teams understood the critical importance of data in their pipelines and the difficulty of finding it. We noticed that AI application engineers are in the same boat, with the availability of quality data playing an overwhelming role in the performance of their products. We are very excited about some of the items on our roadmap. Please [**book a call**]( if you’d like to learn more about what’s coming. Or [**request access**]( if you’d like to get started today! Thanks for reading. We hope to see you around. Jack and Jacob [saldor.com](

Affil.ai

affil-ai

https://affil.ai/

San Francisco, CA, USA

Affil.ai helps financial companies monitor their affiliate marketing content so that they can expand marketing efforts & grow their revenue without compliance concerns.

AI Financial Affiliate Network

B2B

B2B -> Marketing

S24

Active

Early

https://www.ycombinator.com/companies/affil-ai

https://yc-oss.github.io/api/batches/s24/affil-ai.json

Title: Affil.ai: AI Affiliate Network For Financial Brands | Y Combinator URL Source: Markdown Content: ### AI Affiliate Network For Financial Brands We're an affiliate network for financial brands. We help financial companies monitor their affiliate marketing content so that they have full visibility into what they're saying, enabling them to expand marketing efforts & grow their revenue. Affil.ai Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### John Ta, Founder CEO at Affil.ai. Previously: Strategy at Accenture, Growth at Roame (YC S23), M&A IBD at Leerink. Biophysics & Biochemistry major, UPenn ### Vivek Olumbe, Founder CTO @ Affil.ai: Previously at ToughSignal and BLACKIVY. UPenn Comp Sci & Wharton ### Company Launches [### 🤖 Affil: AI Compliance for Financial Affiliate Marketing]( **TLDR;** Financial affiliate programs struggle to ensure affiliates are compliant with brand guidelines, leading to limited growth opportunities. [Affil]( monitors affiliate content, enabling companies to expand marketing efforts and maximize revenue. _What is financial affiliate marketing?_ When you see content by The Points Guy or Forbes that talks about credit cards, checking accounts, etc, they get a kickback from the financial brand for anyone who successfully applies for the product. ### ⚠️ The Problem Financial institutions (e.g., Capital One) or consumer fintech companies (e.g., Acorns) want to grow their affiliate programs but are completely blocked by compliance barriers. However, it’s expensive and time-intensive to do manual affiliate compliance & management. On the flip side, affiliates are also dissatisfied with these financial affiliate programs. The majority have to work with disorganized & outdated middlemen hired by financial brands who take a significant cut of payouts. ### 🛠️ The Solution Our AI compliance software monitors affiliate content so that financial companies know exactly what’s being said about them so that they can expand their marketing efforts: * Track thousands of affiliate pages * Quickly identify compliance violations without needing human intervention * Create support tickets for every issue, so affiliates & brands can keep track of errors We’re also improving the affiliate experience: * Offer higher payouts for conversions (since we have a lower cost structure) * Provide a beautiful user experience & working APIs * Proactively identify compliance risks & offer automated suggestions to minimize risk For example (see below), the AMEX Gold Card underwent significant changes on July 27, 2024, but outdated affiliate content still exists on the internet (that’s not good). We caught this issue that both AMEX and Bankrate missed. The big vision is to become the affiliate network that powers all financial institutions out there. ### 🧍‍♂️Our Team [Vivek]( & [John]( first met at Penn and cofounded the first personal finance nonprofit at 8 different colleges, accumulating 2.5k+ users and $25k+ in funding. Vivek is a full stack developer who worked for several startups shipping new products; John was previously in banking, strategy, and growth roles. After managing affiliate relationships in a prior role & being involved with credit cards for years (fun fact: he has 30 cards), John realized the financial affiliate marketing space was broken. ### 🤝 Our Ask * **Leads:** Know anyone at a consumer finance company or publishers/content creators that write about credit cards/use affiliate links? We’d love an intro! * **Banking:** We are having particular struggles getting in contact with some of the key stakeholders at some of the major banks (think Chase, Wells Fargo, USAA…), so if you can make any warm intros to folks on the consumer banking side, I’ll buy you dinner! Contact: [[email protected]](mailto:[email protected]) ### Other Company Launches ### Affil - AI Affiliate Marketer 🤝 Launch & scale affiliate programs with ease [Read Launch ›](

CodeViz

codeviz

https://www.codeviz.ai/

San Francisco, CA, USA

CodeViz creates interactive maps of codebases to help engineers quickly understand and navigate code. Adopted by engineers at Amazon, Microsoft, and Roblox, CodeViz was founded by two ex-Tesla engineers to tackle the 70% of development time lost to reading code. Download today at https://codeviz.ai

AI agent that builds codebase maps for developers

B2B

B2B -> Productivity

S24

Active

Early

https://www.ycombinator.com/companies/codeviz

https://yc-oss.github.io/api/batches/s24/codeviz.json

Title: CodeViz: AI agent that builds codebase maps for developers | Y Combinator URL Source: Markdown Content: ### AI agent that builds codebase maps for developers CodeViz creates interactive maps of codebases to help engineers quickly understand and navigate code. Adopted by engineers at Amazon, Microsoft, and Roblox, CodeViz was founded by two ex-Tesla engineers to tackle the 75% of development time lost to reading code. Download today at CodeViz Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Liam Prevelige, Founder Building a visual interface for code. Former distributed systems engineer at Tesla. CS + Econ at Dartmouth. ### Will McCall, Founder Building Better Codebase navigation tools. Formerly a Distributed Systems engineer at Tesla. CS + Design at Dartmouth. ### Company Launches [### 📍🗺️ CodeViz - Visual map & GPS for your codebase]( **TL;DR:** CodeViz builds a visual map of your codebase that makes it easy to understand and navigate. Software engineers spend 70% their day reading existing code to gather context - we’re fixing that. **Our Ask:** Try our [VS Code Extension]( [tell us]( what you think, or [set up]( CodeViz for your team! * * * * * * * * * — Hi everyone, we’re [Liam]( and [Will]( and we’re building CodeViz, the dev tool we wish we had at Tesla. From potato cannons to arbitrage bots, we’ve been hacking projects together since we met at Dartmouth 5 years ago. ### ❗The Problem: Like most developers, Will and I were frustrated that 70% of our time was spent reading existing code to gather context, instead of actually writing code. ### 🛠️ Our Solution: Our VS Code extension creates an interactive map of codebases - from high-level system architecture to function calls. Our goal is to give all developers the deep contextual knowledge that’s core to 10x engineers. Engineers from Amazon, Roblox, and Microsoft are already using CodeViz to quickly understand and navigate code. Navigate through the actual flow of your code, rather than file structure. ### Our Ask: Try Codeviz! * Install [CodeViz]( from the VS Code Marketplace and give it a spin in your projects. * Schedule a [chat]( to learn about our business and enterprise packages. If you hear your friends or co-founders complaining about navigating, organizing, or understanding code, please send them our way! Reach out anytime: [[email protected]](mailto:[email protected]) | [[email protected]](mailto:[email protected]) Will and I at Dartmouth’s graduation last year

Mem0

mem0

https://mem0.ai

San Francisco, CA, USA

Mem0 provides a memory layer for LLM applications, enabling them to remember and learn from user interactions over time.

The Memory layer for your AI apps

B2B

B2B -> Infrastructure

S24

Active

Early

https://www.ycombinator.com/companies/mem0

https://yc-oss.github.io/api/batches/s24/mem0.json

Title: Mem0: The Memory layer for your AI apps | Y Combinator URL Source: Markdown Content: ### The Memory layer for your AI apps Mem0 provides a memory layer for LLM applications, enabling them to remember and learn from user interactions over time. Mem0 Founded:2023 Team Size:4 Location:San Francisco ### Active Founders ### Taranjeet Singh, Founder Taranjeet Singh is the co-founder & CEO of Mem0. He joined Khatabook (YC S18) as first growth engineer & quickly transitioned to Senior PM. He began his software engineering career at Paytm (India's Paypal) & witnessed its meteoric rise to become a household name. He built an AI-powered tutoring app that was featured at Google I/O. He co-created EvalAI, an open source Kaggle alternative with Deshraj & grew it to 1.6K GitHub stars. He also created the first GPT app store, scaling it to 1M+ users. ### Deshraj Yadav, Founder Deshraj Yadav is the co-founder and CTO of Mem0. He is broadly interested in the field of Artificial Intelligence and Machine Learning Infrastructure. He led the AI Platform at Tesla Autopilot, which enabled large-scale training, model evaluation, monitoring, and observability for Tesla's full self-driving development. Prior to that, Deshraj created EvalAI, an open source ML platform, as his master's thesis at Georgia Tech. He has published papers at top AI conferences like CVPR, ECCV, AAAI etc. ### Company Launches [### Mem0 - Open Source Memory Layer for AI Apps]( **TL;DR** [Mem0]( is an open-source memory layer for AI applications. It solves the problem of stateless LLMs by efficiently storing and retrieving user interactions, enabling personalized AI experiences that improve over time. Our hybrid datastore architecture combines graph, vector, and key-value stores to make AI apps personalized and cost-efficient. Watch our explainer video [here]( — Hey everyone! We're Taranjeet and Deshraj, and we built [Mem0]( to solve a big problem we faced with LLMs while building Embedchain (an open-source RAG framework with 2M+ downloads). LLMs don’t have memory, so they forget everything after each session. This leads to inefficient and repetitive interactions, making it hard to create personalized AI experiences. Think about having to repeat your preferences over and over again, and how frustrating that gets! Mem0 changes that. ❌ **The Problem** LLMs are stateless—they don’t remember anything between sessions. Every time you interact with them, you have to provide the same context, which gets repetitive and wastes computational resources. This makes AI apps less useful and personalized over time. ✨ **Our Solution** Mem0 adds a **memory layer** to AI applications, making them stateful which allows them to store and recall user interactions, preferences, and relevant context. This way, AI apps evolve with every interaction, delivering more personalized and relevant responses without needing large context blocks in each prompt. To make this possible, we needed to create a system that could efficiently manage and retrieve all the relevant information AI apps collect over time. That’s where Mem0’s **hybrid datastore architecture** comes in, making AI smarter and more efficient as it learns. ⚙️ **How it works** [Mem0]( employs a hybrid datastore architecture that combines graph, vector, and key-value stores to store and manage memories effectively. Here’s the breakdown: * **Adding memories:** When you use Mem0 with your AI App, it automatically detects and stores the important parts of your messages or interactions. * **Organizing information:** Mem0 categorizes memories in three ways: * **Key-value stores** for quick access to structured data (facts, preferences). * **Graph stores** for understanding relationships (like people, places, objects). * **Vector stores** for capturing the overall meaning and context of conversations, allowing AI apps to find similar memories later. * **Retrieving memories:** When an input query is received, Mem0 searches for and retrieves relevant memories using a combination of graph traversal techniques, vector similarity, and key-value lookups. It prioritizes the most important, relevant, and recent information, ensuring that the AI has the right context, no matter how much memory is stored. Watch this video for a demo of our playground in action [here]( 🙏 **Our Asks** * Try out [Mem0]( We guarantee that your users will have personalized interactions after adding Mem0 in your AI apps. Moreover, you will save on your LLM costs. * If you are looking for a memory provider for your AI app and want to schedule a demo, please feel free to block some time on [my calendar]( or email me at [[email protected]](mailto:[email protected]) * Checkout our platform and open-source offering and give feedback: * Platform: [ * GitHub: [ ### Selected answers from Mem0's original YC application for the S24 Batch #### What is your company going to make? Please describe your product and what it does or will do. We are building self-improving memory for LLM apps that enables seamless personalization for end-users. Our product offers APIs that allow developers to store and manage individual user preferences in a centralized layer. This smart memory continuously learns from user interactions, ensuring preferences are consistently applied no matter which LLM the user engages with. It provides a personalized experience across different AI platforms and applications. By offering our smart memory as a service, we empower developers to integrate advanced personalization capabilities into their products, significantly reducing complexity.

Merlin AI

merlin-ai

https://www.merlinai.co

Los Angeles, CA, USA

Merlin is an AI-powered ERP for construction companies. We provide construction companies with a single, AI-powered platform to manage their entire business operations. Our platform unifies project management, materials management, finance, estimating, and CRM—traditionally scattered across multiple systems—into one cohesive solution. For the first time, construction firms can automate and oversee every aspect of their projects in a single, intuitive system. Today, construction companies pay for 8-17 different tools - that we eliminate the need for. We reduce - Manual work by 60% - Cut software costs by $10,000 monthly - Complete projects 3x faster. We're witnessing a perfect storm of factors: decades of stagnation in traditional ERP systems, a multi-billion-dollar industry built on misaligned incentives, and the transformative potential of AI. This confluence creates a rare cyclical opportunity for market upheaval. The last major shift in this space occurred in the 1990s with the advent of digital systems, and now, 30 years later, we're on the cusp of another seismic change. Customers are crying out for innovation and as AI reshapes the landscape, early customer signals indicate an eagerness for smarter, more efficient systems. It's a unique window of opportunity that opens perhaps once in a multi-decade chance to redefine an entire industry's operational backbone.

AI powered ERP for Construction

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/merlin-ai

https://yc-oss.github.io/api/batches/s24/merlin-ai.json

Title: Merlin AI: AI powered ERP for Construction | Y Combinator URL Source: Markdown Content: ### AI powered ERP for Construction Merlin is an AI-powered ERP for construction companies. We provide construction companies with a single, AI-powered platform to manage their entire business operations. Our platform unifies project management, materials management, finance, estimating, and CRM—traditionally scattered across multiple systems—into one cohesive solution. For the first time, construction firms can automate and oversee every aspect of their projects in a single, intuitive system. Today, construction companies pay for 8-17 different tools - that we eliminate the need for. We reduce - Manual work by 60% - Cut software costs by $10,000 monthly - Complete projects 3x faster. We're witnessing a perfect storm of factors: decades of stagnation in traditional ERP systems, a multi-billion-dollar industry built on misaligned incentives, and the transformative potential of AI. This confluence creates a rare cyclical opportunity for market upheaval. The last major shift in this space occurred in the 1990s with the advent of digital systems, and now, 30 years later, we're on the cusp of another seismic change. Customers are crying out for innovation and as AI reshapes the landscape, early customer signals indicate an eagerness for smarter, more efficient systems. It's a unique window of opportunity that opens perhaps once in a multi-decade chance to redefine an entire industry's operational backbone. Merlin AI Founded:2024 Team Size:6 Location:Los Angeles, CA ### Active Founders ### Sneha Kumari, Founder Construction and supply chain expert with 12+ years in prefab, modular, and manufacturing. Implemented multi-million dollar ERP systems and industry best practices across multiple firms and found myself incredibly frustrated and tired from the experience. Founded Merlin to create the ERP solution I always wished for - one that works for construction and manufacturing. Passionate about all things construction and circular economy. ### Ayantika Dey, Founder IIT Computer Science grad. Early engineer at Rippling. Now tackling construction industry pain points head-on with Merlin. ### Kumar Siddhartha, Founder CS grad from BITS Pilani. Spent the last 12 years writing code and building product at startups. 2nd engineer at Swiggy. Co-founded Index ( acquired 2021). Tackling the ERP mess in US construction with Merlin. ### Company Launches [### MerlinAI: AI-powered ERP for construction 🏗️]( **Tl;dr** - We offer construction companies a single, AI-powered platform to streamline their entire business operations. From sales, estimating, project management, and materials management to finance, estimating, and CRM, our platform unifies these essential functions—traditionally spread across multiple systems—into one cohesive solution. * * * — 👷‍♂️👷‍♀️🔧🚧👷‍♂️ Hi everyone, This is Sneha, KS, and Ayantika from Team MerlinAI! [KS]( a BITS Pilani grad and successful startup co-founder, and [Ayantika]( an IIT alumna and tech expert, bring extensive coding and product development expertise together, bringing about over a decade of technical expertise. [Sneha]( with over a decade of experience in supply chain management and lean manufacturing within the construction industry, has lived the pain of using ERPs in this industry. 🚧 **Problem**: Construction companies today juggle 8-17 different software tools—Merlin AI eliminates that complexity. For the first time, firms can automate and manage every aspect of their projects within a single, intuitive system. 🛠️ **Solution**: Merlin AI empowers modular and manufactured home builders with an advanced, AI-driven platform that integrates estimating, materials management, finance, project management, and CRM. Say goodbye to fragmented systems and hello to streamlined efficiency. [ 📌🔨Merlin helps companies reduce * Manual work by 60% * Cut software costs by $10,000 monthly * Complete construction projects 3x faster. 💰💰 **Opportunity**: We're witnessing a perfect storm of factors: decades of stagnation in traditional ERP systems, a multi-billion-dollar industry built on misaligned incentives, and the transformative potential of AI. This confluence creates a rare cyclical opportunity for market upheaval. The last major shift in this space occurred in the 1990s with the advent of digital systems, and now, 30 years later, we're on the cusp of another seismic change. Customers are crying out for innovation, and as AI reshapes the landscape, early customer signals indicate an eagerness for smarter, more efficient systems. It's a unique window of opportunity that opens perhaps once in a multi-decade chance to redefine an entire industry's operational backbone. ✨ **Our Ask** Leads: If you know of a construction company, offsite manufacturer, factory, builder, developer, or design firm wanting to improve their construction integration and project management, please send them to [[email protected]]( 📝 **Mention Us**: If you like what you see, share the love! Mention us in Slack groups, social media, or any community where construction companies, offsite builders, or startup founders gather.

Sensei

sensei

https://senseirobotics.com

San Francisco, CA, USA

Sensei helps robotics companies scale and outsource their training data collection. Our hardware platform enables the collection of human-demonstration data at a tenth of the cost and twice the speed of current teleop approaches. Our software platform acts like Scale AI for robotics data: a large network of paid human operators use our low-cost collection platform to fulfill data-generation requests.

Robotic Training Data at Scale

Industrials

Industrials -> Manufacturing and Robotics

S24

Active

Early

https://www.ycombinator.com/companies/sensei

https://yc-oss.github.io/api/batches/s24/sensei.json

Title: Sensei: Robotic Training Data at Scale | Y Combinator URL Source: Markdown Content: **tl;dr:** [Sensei]( is the Scale AI for robotics training data. Our platform collects human demonstration data at a tenth of the cost and twice the speed of current approaches. **Our Team** ------------ We are MIT engineers who have been friends since undergrad. We worked together at Aurora Flight Sciences, where we ran a DARPA-funded program to develop AI algorithms for autonomous fighter jet combat. [John]( went on to spearhead the reinforcement learning efforts at Aurora, while [Anubhav]( returned to MIT for a PhD focused on robotics, control theory, and machine learning (but dropped out to found Sensei!). We have spent our careers working at the frontiers of robotics and AI. With Sensei, we plan to empower the expansion of those frontiers. **The Problem** --------------- **Data scarcity is one of the biggest challenges in developing robotic artificial intelligence.** Innovators in the space are looking to apply the learnings from the Foundation Model Era: that large models trained on huge datasets are incredibly powerful. As researchers, engineers, and a rapidly growing number of commercial efforts race to develop large models for robotic intelligence, it’s become increasingly clear that the availability of sufficient quantities of quality data is the most significant bottleneck. Current solutions are slow, expensive, and fundamentally not scalable. Human demonstrations are the gold standard for training data. Robotics companies currently employ fleets of 5-20 data-collectors. These in-person contractors teleoperate a robot to perform hundreds of demonstrations a day - with tasks ranging from clothes folding, to bin sorting, to dishwasher loading. **This is not the right long-term solution:** * Large-scale collection is prohibitively expensive: a teleop setup requires a physical robot, often with a cost of $40,000+ per operational platform. * Low-quality data: restriction to labs/office spaces makes it hard to collect varied and realistic demonstrations * Slow data collection rates: non-intuitive interfaces, cumbersome equipment, and faulty hardware lead to slow demonstrations on platforms that break often. These drawbacks make it impossible to scale quality training data collection to the quantities needed for robotics. **Our Solution** ---------------- We design and manufacture low-cost and easy-to-use devices for collecting human demonstration data. Our platform costs less than $300 and can be used by anyone to collect high-quality training data. As seen in a clothes folding task that demonstrates our research prototype, the operator is equipped with a sensorized exoskeleton arm that closely matches the natural human range of motion. The intuitive design, coupled with a suite of angle, vision, and inertial sensors, enables the rapid collection of highly accurate visuo-spatial state information. Our main advantage is that an operator can easily set up and equip the platform, leading to an unprecedented ability to generate quality data at scale. Demonstrations can be performed for a broad range of common tasks, in a nearly infinite number of diverse settings, in as many different ways as human behavior is varied. In order to maximally utilize a fleet of our data-collect platforms, we are building out and operating a network of Senseis— contractors that have been trained to collect high-quality training data using our devices. Our Senseis receive task descriptions from robotics researchers and engineers, and are then paid to collect demonstrations that fulfill the request. **Our combined hardware + software stack represents the first truly scalable solution to generating training data for AI robotics.** **Our Goal** ------------ Within the next five years, we envision tens of thousands of operators, equipped with increasingly sophisticated, portable, and effective data-collection arms, exoskeletons, headwear, and tools. These Senseis will be located all over the world, come from and live in a variety of environments, and all have different takes on what it means to perform tasks “like a human.” This is the best way to power data collection for improved robots, and it is the future we’re building. **Ask** ------- If you’re interested in solving the data scarcity problem in robotics - either for yourself, your company, or your customers - we’d love to hear from you. You can reach us at [[email protected]](mailto:[email protected])

Driftly

driftly

https://driftly.ai

Today, 80% of spending on home healthcare is unrelated to providing care, and most of these tasks can be automated with LLMs. Home health agencies use specialized home health EHRs to manage all aspects of their business--these systems are slow-moving and don't interoperate with third party tools. We're building a new EHR from the ground up. Our first feature is clinician scheduling, which is the biggest pain point for most agencies today. Our software cuts clinician travel time in half, eliminating 10% of total costs and more than doubling profitability.

AI-native EHR for home health agencies

B2B

B2B -> Operations

S24

Active

Early

https://www.ycombinator.com/companies/driftly

https://yc-oss.github.io/api/batches/s24/driftly.json

Title: Driftly: AI-native EHR for home health agencies | Y Combinator URL Source: Markdown Content: ### AI-native EHR for home health agencies Driftly is an AI-native EHR for home health agencies. 80% of the $133B spent every year by home health agencies is unrelated to providing care, and most of this can be eliminated with LLMs. Our first feature is clinician scheduling, where our tool cuts travel time in half, saving 10% of total costs. Driftly Founded:2024 Team Size:2 Location: ### Active Founders ### Toby Hirsch, Founder Toby is the co-founder and CEO of Driftly. Before Driftly, Toby worked as a quantitative trader at Jane Street and at Yale New Haven Health using machine learning to analyze EEG signal for seizure detection and classification. He graduated from Penn's M&T program with degrees in computer science and statistics. ### Victor Lin, Founder Victor is the co-founder and CTO of Driftly. Victor has a background in software and ML, formerly working at Jane Street, Figma, and Bytedance. He formerly founded Umpcast, a scheduling solution for baseball leagues that is used by little leagues across the bay. ### Company Launches [### Driftly: Scheduling automation for home health agencies]( Hey YC – we’re Toby and Victor, and we’re building [Driftly]( to make home health agencies more efficient. **The Problem: Home health clinicians drive too much** ------------------------------------------------------ Clinicians spend 30% of their time driving, while agencies are struggling to meet demand. 75% of referrals to home health are rejected, and with the 85+ population doubling in the next decade, this is only getting worse. Incumbent software is not built for scheduling–instead, agencies rely on Excel and printed territory maps. **The Solution: Route optimization** ------------------------------------ Driftly makes home health clinicians 10-15% more productive by optimizing their routes to cut driving time in half. With Driftly, schedulers can make efficient decisions for a patient’s entire plan of care in just a few clicks. This simple change cuts clinician travel time by 30-50%, allowing agencies to see more patients and increase revenue by 10-15%. **The Team** ------------ Driftly is Toby and Victor’s third startup together. In high school, Victor started Umpcast, a scheduling tool for little leagues that is still in use today across the Bay Area, and Toby was the first hire. Then, they built Probook, a dispatching tool for HVAC and plumbing companies, before leaving to pursue Driftly. [Toby]( (right) previously worked at Yale New Haven Health, where he used machine learning to analyze EEG signals for seizure detection and classification, and as a quantitative trader at Jane Street. He graduated from Penn's M&T program with degrees in computer science and statistics. [Victor]( (left) previously worked at Jane Street, Figma, and Bytedance. He was also involved in the Neo startup community as a Neo Scholar and working at [Relay.app]( He graduated from UCLA with a bachelors in computer science.

SchemeFlow

schemeflow

https://www.schemeflow.com

San Francisco, CA, USA

SchemeFlow uses AI to automate government approvals for construction projects. Two brothers + high school friend, with great combo of skills: Mayor’s Chief of Staff, Arup construction engineer + AI software engineer. Using LLMs to generate reports in minutes rather than months, and to scrape + parse complex local regulations. Already generated over 10,000 pages of technical reports for over 400 construction projects, from car parks to hotels, housing to office buildings, with 97% time savings. Started in the UK with 10x paid pilots with engineering companies such as Stantec and Royal Haskoning, now pivoted to focus on the US. $9B US opportunity (143k consultants x $64k/yr report-writing work), $23B worldwide. www.schemeflow.com

Automating Government Approvals for Construction Projects

Real Estate and Construction

Real Estate and Construction

S24

Active

Early

https://www.ycombinator.com/companies/schemeflow

https://yc-oss.github.io/api/batches/s24/schemeflow.json

Title: SchemeFlow: Automating Government Approvals for Construction Projects | Y Combinator URL Source: Markdown Content: ### Automating Government Approvals for Construction Projects SchemeFlow uses AI to automate government approvals for construction projects. Two brothers + high school friend, with great combo of skills: Mayor’s Chief of Staff, Arup construction engineer + AI software engineer. Using LLMs to generate reports in minutes rather than months, and to scrape + parse complex local regulations. Already generated over 10,000 pages of technical reports for over 400 construction projects, from car parks to hotels, housing to office buildings, with 97% time savings. Started in the UK with 10x paid pilots with engineering companies such as Stantec and Royal Haskoning, now pivoted to focus on the US. $9B US opportunity (143k consultants x $64k/yr report-writing work), $23B worldwide. SchemeFlow Founded:2023 Team Size:4 Location:San Francisco ### Active Founders ### Andrew Browning, CEO & Co-Founder CEO & Co-Founder: Automating Government Approvals for Construction Projects || Experience: Chief of Staff, Mayor's Office in Birmingham, UK || Consultant, Boston Consulting Group || Starred First, St. John's College, University of Cambridge || Founder & Chair of Trustees: King Lear Arts: Active Participation in the Arts for Seniors ### Jonny Browning, Founder CTO & Founder @ SchemeFlow | Automating Government Approvals for Construction Projects ### James Griffith, Founder COO & Co-Founder of SchemeFlow. Automating Government Approvals for Construction Projects. Ex-Arup Engineer, licensed in California, UK & Europe. University of Cambridge Engineering - First Class. ### Company Launches [### SchemeFlow - AI vs. NIMBYs & slow construction projects]( [SchemeFlow uses generative AI to draft the technical reports required for construction projects]( * Getting housing and infrastructure built takes FAR TOO LONG in the developed world today. YEARS are taken up in approvals and technical work before work on site can begin. * Environmental impact assessments, traffic impact studies, acoustics, geotechnical, flood risk reports, applications for federal grant funding. The list goes on… * Our technology drafts these reports in minutes rather than months, allowing technical professionals to focus on the big picture. * Helping to draft these reports quickly and accurately reduces the attack surface for NIMBYs & opponents of projects to slow down or stop projects from getting built. * We HAVE to tackle this problem to restore prosperity and dynamism to our towns and cities. LET'S GET BUILDING! _Elon vs. the environmental technical reports for the Tesla Gigafactory in Berlin_ **WHO ARE WE?** We’re two brothers and a friend from high school in Birmingham in the UK, who started SchemeFlow because of their frustration with how long it takes to get construction projects underway. * [Andrew Browning]( (CEO): Chief of Staff to the Mayor for four years, helping secure a 10x increase in capital funding for transport infrastructure projects. Hugely frustrated with the speed of local projects like the Dudley Metro transit project (first proposal 1984, construction 2020, completion date 2025!) and the paperwork required to get construction started. * [Jonny Browning]( (CTO): Software engineer specialising in the application of AI / ML in existing industries. Previously at Red Hat and Datatonic (European Data & ML consultancy) working on ML platforms for major UK companies such as Vodafone and BT. * [James Griffith]( (COO): Structural engineer in the Advanced Technology Group at engineering firm Arup, licensed in the UK and California. Designed the tallest tower of the Sagrada Familia in Barcelona (the tallest church in the world, accelerating construction speed by 10x) and led the earthquake-resistant glass rebuild of the iconic Museum of Anthropology in Vancouver, BC (making it survive a "once in a millennia" earthquake). _Start a company with your brother: good idea? We’ll find out soon enough..._ **OUR TECH** **Structured Report Generation Platform** * We allow technical consultants to quickly draft these reports in minutes rather than months. * Insert technical analysis quickly and easily into their reports, removing the need for complex third-party software tools for many projects (e.g., GIS platforms). * Access via web app or in Microsoft Word (which is where these consultants live!). **Combination of LLMs and Structured Technical Data** * LLMs to generate a first version, with the ability to quickly tweak and redraft. * Use LLMs on top of technical data & analysis, which is known to be accurate. * Tight control of LLMs & prompts (low temperature) to prevent hallucination. [Read this post]( if you’re interested in what we’ve found using LLMs for technical text generation. _Future of document review: Click, click, click, and choose the version you like best?_ **Add New Structured Reports Quickly and Easily** * Consultants can use their own existing templates, or we can work with them to generate and enhance templates based on previous reports. * Internal tagging tools to speed up our onboarding of new reports and disciplines to the SchemeFlow platform **ASK OF YOU** * **Book a demo** [here]( if you’re an engineer or built environment professional * **Forward this as a link** to an engineer or built environment professional you know * **Follow us** on [LinkedIn!]( #### YC Sign Photo #### Company Photo

Passage

passage

https://www.trypassage.com/

San Francisco, CA, USA

Passage is a co-pilot for the customs brokering space. Today, companies employ entire teams of customs brokers and entry writers who spend hours keying in data and corresponding with clients. With our co-pilot, we plan to streamline this tedious data entry work, and email correspondence, making compliance teams significantly more efficient

Clear your goods in minutes, with AI

B2B

B2B -> Supply Chain and Logistics

S24

Active

Early

https://www.ycombinator.com/companies/passage

https://yc-oss.github.io/api/batches/s24/passage.json

Title: Passage: Clear your goods in minutes, with AI | Y Combinator URL Source: Markdown Content: ### Clear your goods in minutes, with AI Passage is a co-pilot for the customs brokering space. Today, companies employ entire teams of customs brokers and entry writers who spend hours keying in data and corresponding with clients. With our co-pilot, we plan to streamline this tedious data entry work, and email correspondence, making compliance teams significantly more efficient Passage Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Terrell Marshall, Founder Terrell Marshall is the co-founder and CEO of Passage. Before Passage, Terrell worked on AI products with over 1 billion users at Google, like Google Photos, Google Payments, and Hotels Search, and as a software engineer at Workday. Before that, Terrell studied computer science where he met his co-founder Aditya Nadkarni and created a patent for an AI Physical Therapy Platform ### Aditya Nadkarni, Founder Aditya Nadkarni is the co-founder and CTO of Passage. Together with his co-founder, Terrell Marshall, they make customs and logistics workflows more efficient and better through the application of new technologies. Aditya spearheads the development of their platform, drawing from his experience at Flexport's AI and document processing team, and Artera, a start-up streamlining doctor-patient communication. ### Company Launches [### 🧭 Passage - Clearing goods just got easier]( **tldr:** [Passage]( is an AI co-pilot that helps businesses clear their imported goods through customs. **🙏 Our Ask** -------------- * **Do you (or someone you know) work in logistics?** Whether you import goods, move goods (e.g., freight forwarders), or clear goods (e.g., customs brokerages), let us buy you a coffee and book some time with us [here!]( * Just want to give Passage a spin? Try it [here]( for a limited time! **⛔️ The Problem** ------------------ * Every company that brings goods into the US (e.g., Apple, IKEA) needs to clear it through customs * The problem is that clearing customs is super complicated. It has tons of manual work, and if a business screws it up, they face thousands in penalties * This leads to businesses spending hundreds of thousands on customs services, only to spend hours on manual data entry and searching complex regulations **⚙️ The Solution** ------------------- * To tackle this, we’ve built a co-pilot to make clearing customs easier than ever for businesses * Have teams of people manually keying in data? We’ll help streamline the tedious data entry from shipping documents, so your team can spend time on the things that matter * Worried about submitting incorrect data to the government? We’ll check that the data is consistent across all your shipping documents and proactively flag if there’s anything that doesn’t line up * Struggle to understand complex regulations or classify products? We plug directly into key government resources like the code of federal regulations and product tariff schedule so you can clear your goods without the fear of huge penalties **🧑‍🤝‍🧑 The Team** --------------------- [Aditya]( and I ([Terrell]( are friends from college and even have a patent for an AI physical therapy platform. Since college, we’ve moved in together as roommates in San Francisco and have worked on AI products at Flexport and Google. Aditya helped build the AI document processing platform at Flexport (which is how we first learned of the many issues in the logistics industry), and Terrell shipped AI editing features used by Google Photos’ +1B user base. Above is a photo of us after skydiving :) 🪂 #### Company Photo

Diode Computers, Inc.

diode-computers-inc

https://diode.computer

New York, NY, USA

Diode automates hardware design, like hiring a senior electronic engineer for your projects. We use generative AI models to create custom printed circuit boards in minutes, not weeks.

Automate circuit board design using AI

B2B

B2B -> Engineering, Product and Design

S24

Active

Early

https://www.ycombinator.com/companies/diode-computers-inc

https://yc-oss.github.io/api/batches/s24/diode-computers-inc.json

Title: Diode Computers, Inc.: Automate circuit board design using AI | Y Combinator URL Source: Markdown Content: ### Automate circuit board design using AI Diode automates hardware design, like hiring a senior electronic engineer for your projects. We use generative AI models to create custom printed circuit boards in minutes, not weeks. Diode Computers, Inc. Founded:2024 Team Size:2 Location:New York ### Active Founders ### Davide Asnaghi, Founder Davide is the Co-Founder and CEO @ Diode Computers, Inc. - 6 Years of experience building hardware at big tech (Special Projects Group @ Apple working on robotics) and medtech startups (Butterfly Network, Chromatic). - B.Sc. Engineering and Robotics, doing research in Hong Kong and Shenzhen, hardware capital of the world - M.Eng. At Berkeley: designed open source hand prosthesis hardware ### Lenny Khazan, Founder Lenny is the Co-Founder and CTO @ Diode Computers, Inc. - B.S. in Computer Science from Yale University - Attended Apple’s WWDC conference on selective student scholarship: 2012, 2013, 2016 (age 13, 14, 17) - Software + firmware engineering at software startups (Snackpass YC W18) and hard-tech (Butterfly Network, Chromatic) ### Company Launches [### Diode Computers, Inc. — Circuit boards as a service]( Ciao Y Combinator! We’re [Davide]( and [Lenny]( from [Diode Computers, Inc.]( **TL;DR: Diode is an end-to-end platform for circuit board design. With the help of AI, we can finish a months-long design project in days, and then ship working boards straight to your door.** **Try out our designer at **[** or shop generated boards at **[** [ Problem ------- Designing printed circuit boards (PCBs) is really hard. Existing tools provide little opportunity for design reuse, so most projects are started from scratch. Since there’s no good way to test your design, your best bet is to just order them and hope for the best. Thought you were done? Nope — ordering PCBs is a pain too. Now you need to source the components, ship them across the globe to your manufacturer, and constantly be on the lookout for part shortages and replacements. It’s a full-time job. Solution -------- Diode solves this by owning the entire PCB design process end-to-end. It starts with a core of composable, reusable modules that users can stitch together to build their boards. When new use cases come up, our AI generator creates new modules in seconds instead of days. And with rigorously-tested modules instead of de novo designs, we can be confident that each board works first try. We then take the user’s design and automatically translate it into manufacturing requirements to send to the factory. We stockpile all of the necessary components to remove the burden of supply chain management from users and amortize the cost of parts across all designs. The result is an all-in-one platform for hardware design: describe what you need and get working boards delivered to your door. Why --- In a past life, we designed hardware at companies like Apple, and we encountered a problem: PCBs never worked on the first try. At best, they required micro-soldering surgery under a microscope and, at worst, a whole new board. These issues were driving us mad, so we started a company to fix it. We want to make hardware design work more like software. Fast iteration cycles, reusable components with clean composition, and one-click ordering to build your products fast and cheap. Our vision is to turn PCB design into a software problem. **P.S.** Want a customized NFC business card with your logo? Get your own: [ #### YC Sign Photo ### Hear from the founders #### How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?) In a past life, we designed hardware at companies like Apple, and we encountered a problem: PCBs never worked on the first try. At best they required micro-soldering surgery under a microscope, or at worst, a whole new board. These issues were driving us mad, so we started a company to fix it. #### What is your long-term vision? If you truly succeed, what will be different about the world? We want to make hardware design work more like software. Fast iteration cycles, a compiler to yell at you when you make a mistake, and reusable components with clean composition. Today, we’re starting with customizable, build-to-order dev boards for components like the NVIDIA Jetson Nano.

SureBright

surebright

https://surebright.com/

New York, NY, USA; Remote

SureBright embeds AI-powered product warranties in the consumer journey in under 10 minutes, empowering North American merchants to generate additional revenue and boost customer loyalty at zero cost. Think of it as the Stripe of embedded warranties enabling any merchant to offer an Apple Care-style warranty program for every product ranging from smartphones to mattresses. SMBs (< $100M sales) cannot offer warranties due to insurers' high premium requirements. Larger merchants face prolonged onboarding times of over six months and need to build internal tools and support teams to manage warranty operations. SureBright eliminates these barriers, enabling merchants of any size to offer warranties in minutes while reducing the cost to zero. Supporting all sales channels, including major e-commerce platforms, websites, marketplace sales, and brick-and-mortar, SureBright automates every step of the warranty distribution process. By embedding warranties at seven different points in the consumer journey, we achieve over 5x the industry warranty conversion rates. Founded by ex-Amazon and ex-InsurTech Unicorn, we are poised to expand the Product Warranty market by $45 billion.

Offer your own Apple Care-like warranty program in 10 minutes!

B2B

B2B -> Retail

S24

Active

Early

https://www.ycombinator.com/companies/surebright

https://yc-oss.github.io/api/batches/s24/surebright.json

Title: SureBright: Apple Care-like warranty program for every retailer | Y Combinator URL Source: Markdown Content: ### Apple Care-like warranty program for every retailer SureBright lets e-commerce merchants embed "Apple Care” like product warranties into their checkout flows in less than 10 minutes. Our AI-powered warranty platform generates incremental revenue and boosts customer loyalty at zero cost to the merchant. We earn a margin on each policy sold. Currently, we’re live with 60 merchants generating $22K in monthly revenue. We’ve booked an additional $50K of monthly revenue in signed contracts. Sanket led engineering at PolicyBazar, an insurance tech company that IPOed and is valued at over $8B. Manish worked at Amazon and has built $100M ARR products, handling over 180B requests per day. Until now, finding the right policy for any product and pricing it at the right price wasn’t possible at scale. We automated this process using LLM and have unlocked a $50B+ opportunity. Each month, over 5 million consumers view SureBright warranty offers. SureBright Founded:2022 Team Size:7 Location:New York ### Active Founders ### Manish Chauhan, Founder Manish is the CEO and co-founder of SureBright. He holds an MBA from INSEAD (France) and an MMath (CS) from the University of Waterloo, where he researched and published on low-latency real-time systems. He has previously worked at Amazon and startups like Applyboard, where he built products handling over 180 billion requests per day, contributing to $100M+ in ARR. ### Sanket Munjal, Co-Founder CTO & Co-Founder @ SureBright. Led Engineering at an InsurTech, PolicyBazaar.com that went public (8Bn USD). 2x Founder. Ex-Philips, Ex-Samsung ### Company Launches [### SureBright: Offer your own Apple Care-like warranty program in 10 minutes!]( **tl;dr: Embedding these insurance products generates additional revenue, protects customers, and boosts customer loyalty at zero cost. And it’s as easy as signing up for Facebook if you still remember that!** 😉 — [**_Talk to us._**]( _—_ Hello, everyone! We are Manish and Sanket (right to left), the team behind SureBright. We started our professional journey together with our first job at Samsung Research Institute 14 years ago. [**Manish**]( **(CEO)**, a graduate of the University of Waterloo (MMath, CS) and INSEAD (MBA), has developed products for industry giants like Amazon and serves on the board of the InsurTech Association. [**Sanket**]( **(CTO)** has led engineering departments at insurtechs such as Policy Bazaar, which has become a household name in India. Together, we bring 25 years of combined experience in insurtechs, startups, major tech companies, and business. Warranty and insurance distribution in e-commerce has seen minimal change, with the industry evolving slowly. At Surebright, we're on a mission to revolutionize this by leveraging the latest advancements in generative AI. Our platform enables US and Canadian merchants to seamlessly integrate Apple Care-like warranty programs for any eligible product, from ovens to bikes. We achieve this with nearly zero onboarding and operating costs, allowing merchants to offer these warranties in just minutes across any sales channel. ### **What is the problem? ❌** The cost of offering product warranties is exponentially high for merchants and insurers. As it follows, 1. **Resource-Intensive Integration:** Merchants must create a file transfer-based system to integrate with insurers. Additionally, they need to develop internal tools to apply the correct warranty plans to products, build custom UIs, and manage data collection, sharing, and tracking to meet both insurer and regulatory requirements. This entire process can take **six months to a year to go live**. 2. **Recurring Additional Work:** Maintaining integrations and tools requires dedicated engineering teams. Running the warranty program also requires dedicated time from the operation team. This usually requires **additional hiring**. Below is what the current merchant landscape looks like, ### **How is SureBright Solving these problems? 🎉** SureBright’s **warranty program runs on auto-pilot**, empowering merchants of all sizes and providing the following key benefits: 1. **99% less onboarding time, no onboarding or operating costs**, and no recurring tasks or hiring required. Our end-to-end automated solution ensures you never worry about the warranty program. 2. **Cover any product** under warranty and offer other insurance products, such as Shipping Insurance, without lifting a finger. 3. **Enjoy the highest warranty conversions** in the industry, averaging around 20.2%. **With these benefits, we help merchants boost their bottom line** by up to 12% through warranty margin share. Offering warranties also increases customer trust in the brand, leading to 15.6% more conversions and a 14% increase in Average Order Value. In addition, customers get a better post-purchase experience, resulting in more repeat customers and fewer negative reviews. ### **The “aha” moment? 🎯🧠🧨** We began our journey to help smaller merchants (<$100M in sales) offer warranties. But, within 3-4 months of going live, we began attracting interest from larger merchants. **The “aha” moment came when Insurance companies started utilizing our platform for their customers.** This was a pivotal moment, leading us to develop SureBright as the Stripe for embedded e-commerce insurance products. Today, our customers include a diverse range of merchants and marketplaces with $0.5M to $1B in sales, insurance companies, tech firms involved in consumer product purchase journeys (such as Buy Now Pay Later), warranty management companies, and other insurtechs. You can explore our consumer journey with partners such as [**Blackstone**]( **[**Magnum Bikes**]( **[**WalkingPad**]( Street Cred Capital (BNPL), and FlexShopper (Lease to Own).** ### **How can you help? 🙏** If you know any merchant, marketplace, or company participating in the consumer purchase journey, such as Buy Now Pay Later and Checkout Management Companies, please make the introduction. We and they will both be very thankful! Email: [[email protected]](mailto:[email protected]) or [**_BOOK A CALL DIRECTLY WITH US._**]( ### Selected answers from SureBright's original YC application for the S24 Batch #### Describe what your company does in 50 characters or less. We embed product warranties in consumer journeys

Elayne

elayne

https://www.elayne.com/

Denver, CO, USA

Elayne is a benefit solution that empowers employees and their families with estate planning and automated estate closures. Over 75% of today’s workforce will need to plan for or manage a family member’s death in the next decade, a process that takes over 400 hours across 18 months. This administrative burden causes a $100 billion annual drag on US companies due to lost productivity, absenteeism, and attrition. Our platform leverages GenAI and post-pandemic policy changes to automate financial, legal, and administrative tasks. Our AI assistant auto-populates modules based on uploaded materials, identifies gaps in the estate plan, and provides straightforward recommendations. By approaching this problem as an employee benefit, Elayne addresses the 44% increase in attrition and 76% drop in productivity that follows a family member’s death.

AI estate planner & executor for employees and their families

B2B

B2B -> Human Resources

S24

Active

Early

https://www.ycombinator.com/companies/elayne

https://yc-oss.github.io/api/batches/s24/elayne.json

Title: Elayne: AI family estate planning & settlements after death | Y Combinator URL Source: Markdown Content: ### AI family estate planning & settlements after death Over 75% of today’s workforce will need to plan for or manage a family member’s death in the next decade, a process that takes over 400 hours across 18 months. This administrative burden causes a $100 billion annual drag on US companies due to lost productivity, absenteeism, and attrition. Our platform leverages GenAI and post-pandemic policy changes to automate financial, legal, and administrative tasks. Our AI assistant auto-populates modules based on uploaded materials, identifies gaps in the estate plan, and provides straightforward recommendations. By approaching this problem as an employee benefit, Elayne addresses the 44% increase in attrition and 76% drop in productivity that follows a family member’s death. Elayne Founded:2024 Team Size:2 Location:Denver, CO ### Active Founders ### Adria Ferrier, Founder/CEO Co-founder / CEO at Elayne. Former private equity investor focused on the healthcare sector. Wharton grad. ### Jacob Grafenstein, Founder/CTO Chief Technology Officer at Elayne AI. Previously lead engineer on the Rocket Money (Truebill - YC16) web & mobile platform. Built a YouTube channel to 32,000 subscribers. Frequently indie hacking. ### Company Launches [### Elayne - AI estate planner & executor for employees and their families]( Hey everyone, we’re [Adria]( and [Graf]( - cofounders of [Elayne]( 👋 TL;DR ----- Imagine losing a family member. You're grieving and overwhelmed by tasks like funeral planning, managing bills, insurance, probate, and closing accounts—all while trying to process your emotions and return to work. The last thing you need is 400 hours of estate settlement tasks looming over you for the next 16-20 months, making it challenging to perform at work and nearly impossible to take care of your mental health. We solve this problem for both you and your employer. **_Ask:_** Is your HR team currently reviewing benefits for 2025? Introduce us! Email [[email protected]](mailto:[email protected]) or [book a demo here.]( 👫 The Team ----------- [Adria]( (CEO) - Adria is a former private equity investor at Permira and Blackstone and experienced this problem firsthand after losing her mom. Instead of focusing on due diligence at work, she had to spend hours figuring out how to plan a funeral, manage probate, settle bills, and close hundreds of accounts. She left her job within a year of her loss, as many bereaved employees often do. [Graf]( (CTO) - Graf cut his teeth in fintech working for Truebill (YC W16). There, he led the core web and mobile team, managing the implementation of critical experiences such as subscriptions, transactions, and net worth. ❌ The Problem ------------- Employee bereavement costs US businesses $100bn each year. * **A look at the numbers**: For every 1,000 employees _annually:_ * **250 employees** will experience bereavement, 150 will experience a major loss * **418 days** will be spent on affairs settlement by employees * **$608,000** cost to business in lost productivity, absenteeism, and attrition * **No solutions:** there are no practical tools in the market to help HR leadership support their employees in the months following a loss * **Manual and time-consuming**: settling an estate takes an average of 400 hours over 18 months to complete and is currently completely manual 🙇 Our Solution: One Platform for Family Estate Planning and Automated Settlements ---------------------------------------------------------------------------------- **Family Estate-Readiness 👪** The estate-readiness platform provides a novel way for employees and their immediate family members to organize and optimize their estates. * **Visualizations**: Visualize your estate by uploading existing documentation and using our AI to extract information. * **Optimizations**: Use our AI to identify issues that could lead to problems executing the estate. * **Document Vault**: Upload and securely store critical documents that are needed in the event of a death. **Automated Settlements 🕊️** The post-loss platform saves employees 400 hours over 18 months by automating estate and affairs settlement after each loss. * **AI Concierge**: Chat with our AI concierge, who knows all the details about your situation and can execute tasks on your behalf. * **Personalized Roadmap:** Understand at a glance what you have to do, what your loved one planned ahead of time, and what the Elayne AI can do for you. * **Resources:** Access exclusive resources that help you understand probate and other opaque parts of the post-loss process. **For HR Leaders ✨** Knowing what to say and how to support an employee experiencing a loss is hard, and navigating the death of an employee is even harder. We're here to help: * **Immediate Response**: Elayne steps in with an immediate action plan to coordinate urgent needs and communications on behalf of the employee or their family * **Workshops:** Explore the emotional and practical challenges that arise during the estate planning and settlement process and learn strategies to foster resilience * **Dashboard:** Quickly understand utilization rates, hours saved, and cost savings -for both your employees and company. 🙏 Our Ask ---------- * Follow us on [LinkedIn]( and [book a demo]( with our team. * **Share this post!** \-  Do you know anyone in your network who is a CHRO, HR or benefits manager, or head of Total Rewards? Please share this post (or our promo video below!) * Learn more at [Elayne.com]( [

Drillbit

drillbit

https://www.drillbit.com

Austin, TX, USA

Residential contractors are drowning in office work. Drillbit automates all of it. We use AI receptionists and an LLM based quoting tool on top of our own CRM to entirely manage jobs from lead, to quote, to scheduling, to staffing, and even collecting payment.

Automating Office Work for Residential Contractors.

B2B

B2B -> Operations

S24

Active

Early

https://www.ycombinator.com/companies/drillbit

https://yc-oss.github.io/api/batches/s24/drillbit.json

Title: Drillbit: Automating Office Work for Residential Contractors. | Y Combinator URL Source: Markdown Content: ### Automating Office Work for Residential Contractors. Residential contractors are drowning in office work. Drillbit automates all of it. We use AI receptionists and an LLM based quoting tool on top of our own CRM to entirely manage jobs from lead, to quote, to scheduling, to staffing, and even collecting payment. Drillbit Founded:2023 Team Size:2 Location:Austin, TX ### Active Founders ### Alexander Seutin, Founder Alex graduated from Stanford with a degree in CS & AI. After school, he worked as an engineer at SpaceX, Apple, Rivian, and Built Robotics. Alex is passionate about affordable housing, and after years of doing construction on the side, he merged his passions and started Drillbit. His company seeks to optimize and scale home service businesses through automating their operational & administrative workflows. ### Liam Osler, Founder I make powerful and complex systems accessible to underserved industries. At Drillbit we simplify the operations of home service businesses using AI. ### Company Launches [### Drillbit 🧰 Handling office work for residential contractors]( ### **~TLDR~** Residential contractors are drowning in office work that has nothing to do with their core business. [**Drillbit**]( automates all of it. We use AI receptionists and an LLM based quoting tool on top of our own CRM to entirely manage jobs from lead, to quote, to scheduling, to staffing, and finally payment. ### **Problem🚨** Residential contractors provide a critical service to our economy: building and maintaining the supply of housing. Unfortunately, those businesses are still operated in an incredibly manual way. Across the U.S., we spend about $675B/year on simple home maintenance. A significant portion of that cost is attributed to the office work involved with finding, coordinating, and managing those small jobs. Business owners need to (and prefer to) remain focused on the real work instead of non-value-driving efforts. ### **Solution 🏗️** [**Drillbit**]( allows contractors to be entirely hands-off, enabling them to focus 100% of their time on their core business. We have built AI receptionists that can coordinate with businesses and homeowners at every stage in the process. The receptionists then work with our AI estimators to provide quotes. Together, they are able to use our CRM, project management, and payment solutions to store updates and take action throughout the process. ### **Landscape ⛰️** Platforms that have come before us were able to move service business operations from texting and emailing over to CRMs and project management solutions, but they suffer from massive churn rates. Owners we have spoken with have consistently voiced the feeling that they were “babysitting” the tool. All of that “homework” they had to do entirely undermined the value of the solution in their eyes, and so, most users would abandon the platform. Rather than expecting our users to change their behavior - ie. start using technology that they would rather not be interacting with - we have met them where they are by handling the job coordination in a way that feels familiar to them: offering the ability to talk to your “employees” on the phone like you would a real person, and having those “employees” be able to engage with homeowners over the phone like real people would too. ### **Team 👷👷‍♂️** [**Alex**]( (Stanford CS / AI) and [**Liam**]( (WashU CS) first worked together developing robots for industrial construction projects. Alex would go on to spend a few years renovating airstreams and building a house for his family. During that time, Alex became intimately familiar with the pain points in the construction and home service industries and realized how much of a need there was for better, more automated software. At that point, Alex called Liam, and together, they started building [**Drillbit**]( to support the folks maintaining America's housing. ### **Ask 🙏** The platform has been built, and now we are looking to drive revenue. If you know of any business owners in the industry (plumbers, electricians, house cleaners, lawn care providers, landscapers, tilers, roofers, etc) please introduce us at [**[email protected]**](mailto:[email protected]) 🙂 ### **Links 🔗** In the meantime, we’d love it if you followed our journey: * [**Company LinkedIn Page**]( * [**Alex’s LinkedIn Page**]( * [**Liam’s LinkedIn Page**]( & finally, check out our website! [** Demo Video: [** #### YC Sign Photo #### Company Photo ### Hear from the founders #### How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?) Alex and Liam first worked together developing robots for industrial construction projects. Alex would go on to spend a few years renovating airstreams and building a house for his family. During that time, Alex became intimately familiar with the pain points in the construction and home service industries, and realized how much of a need there was for better, more automated software. At that point, Alex called Liam and together they started building Drillbit to support the folks maintaining America's housing. #### What is the core problem you are solving? Why is this a big problem? What made you decide to work on it? Across the United States, we spend $700B/year on maintaining homes. Half the cost of any given home service is attributed to compensating the work it took just to coordinate that job. We believe that can and should be solved for. Owners need to (and prefer to) remain focused on “the real work” instead of non-value driving efforts. Businesses in this industry provide an essential service to our economy - they maintain our supply of housing - but they are operated in an incredibly manual and wasteful way. We solve that with AI receptionists that can coordinate with each of the parties at every stage in the process. The receptionists can then work with our AI estimators to provide quotes, and together they can use our CRM, project management, and payment solutions to store updates and take action throughout the process. #### What is your long-term vision? If you truly succeed, what will be different about the world? There are many elements that contribute to the inflated cost of housing and they impact everyone from the homeless, to renters, to homeowners, and investors. An often-overlooked contributor to this cost is the high opEx in owning and maintaining a home. The most wasteful element of that expenditure is not materials or labor - it is compensating businesses for their administrative work (finding, quoting, managing, and coordinating jobs). My dream is to cut out those inefficiencies by bringing true automation to the home service industry. In doing so, we as a society can take a step towards better, more accessible housing. ### Selected answers from Drillbit's original YC application for the S24 Batch #### Describe what your company does in 50 characters or less. AI powered ERP for Home Service Businesses #### How long have each of you been working on this? How much of that has been full-time? Please explain. Alex: Full-time for 8 months. Liam: Full-time for 2 months. #### What is your company going to make? Please describe your product and what it does or will do. Home service businesses spend half their day finding work and doing administrative tasks. Jobber & ServiceTitan helped by moving businesses from paper to digital, but those platforms still require a huge amount of manual input and babysitting. AI has unlocked the next wave of automation. We already have an instant quoting system that is solving a huge pain point for SMBs: getting work. It does this by expediting the process from lead to quote. Our platform will eventually cut out all of the manual processes involved in operating any service-based business, enabling its personnel to cut out the busywork, "get back to building", and double their revenue. ### YC S24 Application Video

Overlap

overlap

https://www.joinoverlap.com/

Overlap is Perplexity for video & audio content

Consumer

Consumer -> Content

S24

Active

Early

https://www.ycombinator.com/companies/overlap

https://yc-oss.github.io/api/batches/s24/overlap.json

Title: Overlap: Overlap is Perplexity for video & audio content | Y Combinator URL Source: Markdown Content: ### Overlap is Perplexity for video & audio content Overlap is Perplexity for the millions of hours of smart YouTube and interview content out there. Prompt Overlap with any topic and get the perfect clip. We have 7k users growing over 200% month-over-month since launching. ### Jobs at Overlap San Francisco, CA, US $130K - $180K 0.50% - 2.00% Any (new grads ok) Overlap Founded: Team Size:3 Location:San Francisco ### Active Founders ### Jonathan Baer, Founder/CEO CEO & Founder of Overlap ### Buzz Rubenstein, Founder/COO Cofounder & COO of Overlap ### Casey Traina, Founder/CTO CTO & Cofounder at Overlap ### Company Launches [### Overlap - Curating the best podcast moments for you]( Hey everyone, we’re Jonathan, Casey, and Buzz. We’re on a mission to make the podcast experience more personalized and accessible for the 500M global podcast listeners. ❌ **The Problem** ----------------- There are 5 million hours of new podcast content each week, distributed via archaic RSS feeds. The moments that make people love podcasts are often lost in the deluge. Sifting through hours of content to find the best clips is time-consuming and overwhelming. **✨ Our Solution** ------------------ Overlap is the first podcast experience that uses LLMs to curate the best moments in any given episode. It creates personalized playlists of clips based on your interests. Ask about any topic, and it can retrieve the most relevant, high-value clips from a wide range of shows. It makes the podcast universe a joy to explore. **🔥 Community fave features** ------------------------------ **👉 Every day, Overlap curates the best clips from your favorite shows into custom playlists.** **👉 Ask about any topic, and Overlap will take you down a rabbit hole.**  **👉 Share your favorite clips with your friends.** **🙏 Asks:** 1. **Try Overlap:** It's available on the App Store [**here.**]( 2. **Feedback:** Feedback is crucial. We rely on our users to tell us what they think. If you have thoughts, please book a time here: [** 3. **Share it:** If you like it, share it with friends and make it your go-to podcast app. 4. **Follow:** Join us on [**Twitter.**]( \- [Jonathan]( [Buzz]( [Casey](

Lumenary

lumenary

https://lumenary.com

San Francisco, CA, USA

Lumenary automates value-based quality reporting for healthcare providers using AI, unlocking tens of millions of dollars in reimbursements and incentives. With just a few clicks, we unify EHR & claims data, calculate risk-adjusted quality measures, and submit hundreds of different reports to payers, registries, and regulators.

AI-powered quality reporting for healthcare.

Healthcare

Healthcare

S24

Active

Early

https://www.ycombinator.com/companies/lumenary

https://yc-oss.github.io/api/batches/s24/lumenary.json

Title: Lumenary: AI-powered quality reporting for healthcare. | Y Combinator URL Source: Markdown Content: ### AI-powered quality reporting for healthcare. Lumenary automates value-based quality reporting for healthcare providers using AI, unlocking tens of millions of dollars in reimbursements and incentives. With just a few clicks, we unify EHR & claims data, calculate risk-adjusted quality measures, and submit hundreds of different reports to payers, registries, and regulators. Lumenary Founded:2024 Team Size:4 Location:San Francisco ### Active Founders ### Vivek Hazari, Founder, CEO Vivek is the CEO and co-founder of Lumenary. Prior to starting Lumenary, Vivek spent 2 years at Palantir Technologies, where he worked across the commercial business, and later help found and lead a team to aid the development of AIP. He studied Computer Science and Philosophy at Dartmouth College, where he did research on applications of language models. ### Daniel DiPietro, Founder Dan is the co-founder and COO of Lumenary. Previously, Dan worked at Jane Street, where he built capital allocation optimizations used in production to invest billions of dollars of assets, as well as an options tool used to execute \>$100m of trades annually. He has published multiple first-author publications in high-impact venues like NeurIPS and CoRL. Dan moonlights as a botanist and published a best-selling botanical textbook on carnivorous plants while studying at Dartmouth College. ### Ziray Hao, Founder / CPO Ziray is a cofounder at Lumenary. Prior, he designed AI product for Palantir Technologies. ### John McCambridge, Founder John is the co-founder and CTO of Lumenary. As a former high-frequency trading engineer at IMC, John specializes in low-latency, high-throughput trading algorithms and data services. Born in New Zealand and raised in Scotland, he enjoys flying small aircraft (occasionally jumping out of them), building small-scale rockets, exploring vulnerable endpoints and playing poker. ### Company Launches [### Lumenary: Automated quality reporting for healthcare]( tl;dr ----- Each year, hospitals submit hundreds of quality measures to payers, regulators, and registries, which costs each hospital at least 65,000 labor hours annually. [Lumenary]( uses AI to automate this process, **reducing administrative burden** and **unlocking tens of millions in additional revenue** from value-based contracts. Why this, and why us? --------------------- Hi everyone! We’re [Dan]( [Vivek]( [John]( and [Ziray]( the team behind [Lumenary]( We met each other during our first few weeks at Dartmouth and have been best friends ever since. Our team brings together engineering from Palantir, Jane Street, IMC Trading, and Flatiron Health. We’ve built enterprise-grade AI products, a capital allocation system that controls billions, and a data engine that processes petabytes–we _all_ build and deploy software _quickly_ in contexts _where it cannot fail_. Our journey began with a simple desire to provide feedback after a hospital visit and not knowing how to. We began reading about patient experience but soon learned that it’s just one component of a broad, complicated system of healthcare quality management. After speaking with dozens of clinicians and administrators, we began to unravel the complicated web of quality measurement, reporting, and improvement. We learned exactly where and how these measures impact a hospital’s bottom line, but just how time-consuming the process is now. The Problem ----------- “Value-based care” is one of the most important concepts in healthcare today. _Value_ means getting as much _quality_ per dollar as possible, but quality is hard to define and harder to measure. Quality reporting is a massive time and money sink. Health systems employ large teams to manage the process, **costing $4-12 million per hospital annually just in direct labor costs,** not even counting the documentation burden placed upon clinicians themselves. There is enormous complexity in the process: hundreds of quality measures, dozens of submission forms per payer, and overlapping standards and requirements from registries and regulators. **Errors, delays, and suboptimal reporting can cost tens of millions of dollars in fines and missed reimbursements.** Once you know where you stand, improving your scores is equally cumbersome, requiring dozens of data analysts searching for patterns manually. What we do ---------- Lumenary’s Integrated Quality Platform puts a system that requires at least 65,000 hours of labor annually on autopilot. 1. We unify disparate health data (EHR, insurance claims, patient surveys). 2. We do automatic data abstraction over our unified layer to calculate each quality measure according to specific reporting needs and methodology. 3. We deliver and analyze quality data in real-time, enabling health systems to plan improvements and provide point-of-care reminders within the EHR, enhancing adherence to quality measures. 4. We handle submissions to payers, registries, and regulators in all required formats. Our Ask ------- We’re looking to connect with healthcare professionals who could benefit from our product, especially: * Healthcare leadership (CEO/CMO/Chief Quality Officers) * Leadership of small clinics, especially ACOs * Clinicians or unit leaders who are upset with administrative burden * Healthcare data analysts

AI Sell

ai-sell

https://www.aisell.com

San Francisco, CA, USA

We're building AI Sales Associates for eCommerce. Imagine going on nike.com and facetiming an AI Lebron James and getting his advice on which his shoes to buy. Traditional brick-and-mortar retail stores rely on sales associates to provide personalized customer service, guide product discovery and close sales, but it's always been too expensive and unscaleable for online stores. We're building video-based AI agents that are experts in their products and can gently steer customers to a purchase, delighting customers and lifting conversion/AOV. Jerry and Jeffrey are brothers who previously worked together to scale The Coding School to a 6mil+ in revenue. Jerry built software to help run the 50+ Shopify brands that Openstore acquired more profitably and efficiently. He was also an engineer at Flexport and Google. Jeffrey previously worked at DRW as a quant trader, and as an engineer at Facebook, Snowflake and Amazon.

AI Sales Associate for Ecommerce

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/ai-sell

https://yc-oss.github.io/api/batches/s24/ai-sell.json

Title: AI Sell: AI Sales Associate for Ecommerce | Y Combinator URL Source: Markdown Content: ### AI Sales Associate for Ecommerce We're building AI Sales Associates for eCommerce. Imagine going on nike.com and facetiming an AI Lebron James and getting his advice on which his shoes to buy. Traditional brick-and-mortar retail stores rely on sales associates to provide personalized customer service, guide product discovery and close sales, but it's always been too expensive and unscaleable for online stores. We're building video-based AI agents that are experts in their products and can gently steer customers to a purchase, delighting customers and lifting conversion/AOV. Jerry and Jeffrey are brothers who previously worked together to scale The Coding School to a 6mil+ in revenue. Jerry built software to help run the 50+ Shopify brands that Openstore acquired more profitably and efficiently. He was also an engineer at Flexport and Google. Jeffrey previously worked at DRW as a quant trader, and as an engineer at Facebook, Snowflake and Amazon. AI Sell Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Jerry Liu, Cofounder, CEO CEO/Cofounder @ AI Sell. Excited to build on my work at Openstore where I built software to run 50+ Shopify brands more profitably. Also worked on payments at Flexport and ads at Google. Was CTO of The Coding School, a nonprofit that raised 6m+ from Google/IBM/Department of Defense. CS @ Duke. ### Jeffrey Liu, Founder CTO/Cofounder @ AI Sell. Previously Quant Trader @ DRW and SWE @ Snowflake, Meta, AWS. CS @ Duke ### Company Launches [### AI Sell: AI Sales Associates for eCommerce]( **TL:DR** [AI Sell]( helps eCommerce stores increase conversion/AOV with AI sales associates expertly trained in their product catalog and brand voice. The Problem ----------- Traditional brick-and-mortar retail stores have always relied on sales associates to provide personalized customer service, guide product discovery, and close sales, but it's always been too expensive and unscalable for online stores. The Solution ------------ We work with you and your brand to craft AI agents that reflect your unique brand voice and are experts in your product catalog. We use ML models to detect when users are confused, indecisive, or about to drop off and engage them at the right moment to convert them. The Team -------- Jerry and Jeffrey are brothers who have teamed up to build software for eCommerce. Jerry previously built software to run the 50+ Shopify brands that Openstore acquired more profitably and efficiently. Jeffrey did engineering at Facebook/Amazon/Snowflake. They both graduated from Duke with degrees in Computer Science. ### Sign up for a demo today Schedule a demo on [aisell.com]( or reach out to [[email protected]](mailto:[email protected]). Also, if you know any Shopify owners/operators, please share this with them!

Asterisk

asterisk

https://asterisk.so

Asterisk (asterisk.so) is an AI Hacker for Codebases. It automatically finds, verifies, and patches security vulnerabilities in codebases, just like a human security engineer would. Asterisk can find business logic errors with context-aware scanning and automate the full auditing cycle to generate reports with near-zero false positives. Asterisk has autonomously discovered vulnerabilities in Google, Hoppscotch, and others.

AI Hacker for Codebases

B2B

B2B -> Security

S24

Active

Early

https://www.ycombinator.com/companies/asterisk

https://yc-oss.github.io/api/batches/s24/asterisk.json

Title: Asterisk: AI Hacker for Codebases | Y Combinator URL Source: Markdown Content: ### AI Hacker for Codebases Asterisk (asterisk.so) is an AI Hacker for Codebases - it automatically finds, verifies, and patches security vulnerabilities in codebases, just like a human security engineer would. Asterisk can find business logic errors with context-aware scanning and automate the full auditing cycle to generate reports with near-zero false positives. Asterisk has autonomously discovered vulnerabilities in Google, Hoppscotch, and others. ### Latest News Asterisk Founded:2024 Team Size:3 Location: ### Active Founders ### Mufeed VH, CEO / Founder CEO @ Asterisk, Mufeed is an Olympian CTF-er (security hacking competitions) representing India at the WorldSkills Olympiad (🏅), BRICS Skills (🥉), and IndiaSkills (🥇). He has worked on 300+ CVEs in open source. Secured Google, Mastercard, Okta, DELL, Ford, Zilliqa, EFF, Telefonica, etc. via Bug Bounty programs. mufeedvh.com ### Asjid Kalam, COO / Founder COO @ Asterisk, Former security research engineer, and international CTF player. Solving the hardest code security challenges with AI. Asjid has contributed to multiple CVEs and secured companies like Google and Mastercard via bug bounties. ### Vivek R, CTO / Founder CTO @ Asterisk. Previously, Vivek built secure, fault tolerant distributed systems for leading POS validators handling a total stake of $2B+, did IoT protocol optimization research for startups, contributor to GStreamer - the leading multimedia framework used by Google, Apple, etc, helped Fortune 100 companies with their infrastructure strategy and optimization. ### Company Launches [### Asterisk: AI-automated Security team]( **TL;DR**: [Asterisk]( is an AI “hacker” with the ability to find, exploit, and patch security loopholes across your digital assets. It does all of this with no user intervention and generates a report with zero false positives. ([example]( **The Problem** --------------- * Almost 95% of potential vulnerabilities raised by static security tools (SAST) are **false positives**, and security engineers spend a lot of time filtering them out. * Current security tech **cannot detect business logic bugs** - think of vulnerabilities like unauthorized access, privilege escalation, and bugs that would rack up your AWS/API bills. **Solution** ------------ * **Verification:** Asterisk verifies a vulnerability by spinning up a sandbox, running the software being scanned, and attempts to exploit the bug. If Asterisk flags a vulnerability, you can be sure it's real. * **Context-aware Threat Modeling:** Asterisk understands your codebase and can thus emulate a malicious hacker by coming up with attack scenarios, like the recent CrowdStrike incident. **Our Team** ------------ We are [Mufeed]( [Asjid]( and [Vivek]( All of us have a background in security research and systems engineering. **Mufeed**: represented 🇮🇳 at WorldSkills CTF, 🥇 medalist - IndiaSkills CTF, 🥉 medalist - BRICS Skills CTF **Asjid**: 🥈 medalist - IndiaSkills, ex-security research engineer at Emirates National Bank (UAE) **Vivek**: ex-distributed systems/platforms engineer at Chorus One (one of the biggest POS validators) Our team has helped secure Google, Mastercard, Okta, Nvidia, Microsoft, etc. We are also the team behind [Devika,]( the open-source alternative to Devin with over 18K stars on GitHub. **Our Ask** ----------- If you're looking for a **complete security audit** of your digital assets or want to talk about anything security, email us ([[email protected]](mailto:[email protected])) or book a demo ([asterisk.so/demo]( #### YC Sign Photo

Standard Data

standard-data

https://www.standarddata.com/

San Francisco, CA, USA

Standard Data builds LLM connectors to integrate manufacturing data. At Palantir, we saw manufacturers spend hundreds of millions of dollars to solve this problem manually with legacy systems such as SAP and Oracle, but we realized this problem can be automated using LLMs.

Building LLM connectors to integrate manufacturing data.

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/standard-data

https://yc-oss.github.io/api/batches/s24/standard-data.json

Title: Standard Data: Building LLM connectors to integrate manufacturing data. | Y Combinator URL Source: Markdown Content: ### Building LLM connectors to integrate manufacturing data. Standard Data builds LLM connectors to integrate manufacturing data. At Palantir, we saw manufacturers spend hundreds of millions of dollars to solve this problem manually with legacy systems such as SAP and Oracle, but we realized this problem can be automated using LLMs. Standard Data Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### John Robison, Founder Building LLM connectors to integrate manufacturing data. ### Shaashwat Sharma, Founder Helping organizations standardize their data ### Company Launches [### Standard Data - AI Data Analyst for manufacturers]( ### **Intro** Hi, we’re [John]( and [Shaashwat]( We worked together at Palantir in the manufacturing vertical, where John focused on industrial chemicals while Shaashwat worked on commercial defense. Our work on the ground exposed some incredible risks to American advanced manufacturing, particularly around how data is managed (or mismanaged) in an operational context. ### **Problem** * Data is stored in many formats, including various ERPs, PDFs, Word documents, and Excel files, across multiple different systems on the factory floor. * This setup makes it difficult to answer basic questions, and answers are more often found in the heads of subject matter experts rather than in an easily accessible, uniform platform. ### **Solution** Our solution enables manufacturers to: * Auto-integrate their disparate data into a singular knowledge base * Query this knowledge base in natural language to provide visibility across their core processes. * Implement workflows on top of this knowledge base to increase automation and improve overall decision-making. By capturing and leveraging operational knowledge more effectively, this will ensure American manufacturing can continue to expand capacity in the coming decades. ### **Why Now?** * A wave of retirements, coupled with insufficient incoming skilled labor, will open up an estimated [**2.1 million unfilled advanced manufacturing roles by 2030**]( * This knowledge loss necessitates increased efforts to automate processes and manage data more effectively to ensure these factories continue running at full capacity. ### **Curious for More?** Thanks for reading! If you would like to get in touch to ask questions or offer feedback, please email us directly at [**[email protected]**](mailto:[email protected]), or [**schedule a meeting**](

Pulse AI

pulse-ai

https://www.trypulse.ai

San Francisco, CA, USA

Pulse automates procurement with AI to reduce supply chain risk. Our AI agents eliminate the time wasted on fragmented supply chain data in emails, spreadsheets, and ERPs.

Procurement on autopilot

B2B

B2B -> Supply Chain and Logistics

S24

Active

Early

https://www.ycombinator.com/companies/pulse-ai

https://yc-oss.github.io/api/batches/s24/pulse-ai.json

Title: Pulse AI: Procurement on autopilot | Y Combinator URL Source: Markdown Content: ### Procurement on autopilot Pulse automates procurement with AI to reduce supply chain risk. Our AI agents eliminate the time wasted on fragmented supply chain data in emails, spreadsheets, and ERPs. Pulse AI Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Sid Manchkanti, Founder Co-Founder/CEO of Pulse AI (S24). Prev NVIDIA, D.E. Shaw, Berkeley CS ### Ritvik Pandey, Founder Co-Founder/CTO of Pulse AI (S24). Prev Tesla ML, Goldman Sachs, Georgia Tech CS/Math ### Company Launches [### ⚡ Pulse AI - Procurement on autopilot]( We help companies automate procurement workflows, supplier reliability analysis, and ML-based demand forecasting, giving the company more time to prioritize critical and innovative tasks. Employees are saving hours a day and are ~25% more productive on previously tedious data transcription tasks. Hey everyone, we’re Sid and Ritvik — the founders of [Pulse AI]( ❌ **The Problem** ----------------- Legacy ERPs are extremely antiquated, expensive, and slow. However, they are a vital part of nearly every company. When we were working at Tesla and NVIDIA, we saw firsthand how supply chain teams wasted hours of time on inefficiencies + manual transcription tasks that could have easily been solved with better ERP tools. ✅ **The Solution** ------------------ Pulse AI offers intelligent automation tools, like supplier health tracking and onboarding, procurement automation, semantic search engine, planning analytics, and much more. We know how difficult it is to swap out entire ERP systems from the very beginning. This is why our product integrates on top of existing ERPs (SAP, NetSuite, etc.) alongside email, messaging tools, and document suites in under 10 minutes. Using Pulse AI’s software, **employees**: ✔️ Can save hours a day on previously tedious searches and disorganized information. And **companies**: ✔️ Save millions of dollars by identifying and fixing supply chain inefficiencies. ✔️ Automate dozens of back-office supply chain functions, giving the company more time to prioritize critical and innovative tasks. **🧩 Team** ----------- The founding team has deep machine learning experience at Tesla, NVIDIA, D. E. Shaw, and AWS — as well as research experience at world-class AI labs at Berkeley and Georgia Tech. Our team also consists of supply chain advisors who have worked in the industry at Fortune 500 companies for the past 25 years. We are super excited to be working with veterans in supply chain! 🙋**Our Ask** ------------- * Please share this post, as you never know who it may help! Feel free to contact us directly at [[email protected]](mailto:[email protected]) * Follow us on [Linkedin]( and [book a demo with our team here]( * Do you know anyone that deals with supply chain processes? Connect us with global supply chain directors, planners, and/or operations strategists! **_A quick blurb to copy & paste_**_:_ A team of former Tesla, NVIDIA, and AWS employees recently created a [software suite]( to automate manual processes from legacy supply chain systems with AI. Check them out and contact [[email protected]](mailto:[email protected]) for more.

ZeroPath

zeropath

https://zeropath.com

San Francisco, CA, USA

ZeroPath is a developer tool that autonomously detects, verifies, and submits fixes for vulnerabilities in your code. Engineers can use ZeroPath to find security problems that they might only otherwise catch in pentests or from bug bounty researchers.

Automatically find and fix your software vulnerabilities

B2B

B2B -> Security

S24

Active

Early

https://www.ycombinator.com/companies/zeropath

https://yc-oss.github.io/api/batches/s24/zeropath.json

Title: ZeroPath: Automatically find and fix your software vulnerabilities | Y Combinator URL Source: Markdown Content: ### Automatically find and fix your software vulnerabilities ZeroPath is a developer tool that autonomously detects, verifies, and submits fixes for vulnerabilities in your code. Engineers can use ZeroPath to find security problems that they might only otherwise catch in pentests or from bug bounty researchers. ZeroPath Founded:2024 Team Size:4 Location:San Francisco ### Active Founders ### Nathan Hrncirik, Founder CIO at ZeroPath. Security Engineer & 100k+ earned bug bounty hunter. Formerly Red Team at Tesla. ### Raphael Karger, Founder CTO at ZeroPath. Former Google Security Engineer and BishopFox Consultant. ### Etienne Lunetta, Founder COO at ZeroPath. Former Co-Founder at Mevlink, acquired by bloXroute Labs in 2023. ### Dean Valentine, Founder CEO of ZeroPath. Previous exit at Mevlink, acquired by Bloxroute in 2023. ### Company Launches [### 🛡️ ZeroPath - Autonomous vulnerability patching]( 📝 TL;DR: [ZeroPath]( is a GitHub app that detects, verifies, and issues pull requests for security vulnerabilities in your code. ### 💡 Why We Built This: As security engineers and developers, we've seen firsthand how challenging it is for businesses to secure their products. Existing Static Application Security Testing (SAST) tools overwhelm teams with false positives and fail to remediate the real vulnerabilities they do find. We built ZeroPath to allow fast-moving teams to write secure code without having to configure or review a scan dashboard. ### 🛠 How It Works: 1. Executes regular vulnerability scans of your source code 2. Integrates with GitHub to scan pull requests for vulnerabilities 3. Uses advanced LLMs to significantly reduce false positives 4. Generates patches for vulnerabilities identified 5. Automatically issues pull requests to patch them ### 💪 Why ZeroPath: * Integrates with your existing SAST - **removes** **85% of false positives** * Reduces vulnerability remediation time by up to **90%** * Enables natural language modification of patches, just tag @zeropath-ai in PRs (including your own!) * Supports complex, multi-file changes seamlessly * Prevents SAST vendor lock-in ### 👥 Our Team: * [Nathan]( Ex-Tesla Red Team Engineer, $100k+ in bug bounties * [Raphael]( Ex-Google Security Engineer, 12 security researcher hall of fames * [Dean]( & [Etienne]( Founders of Mevlink (acquired by Bloxroute in 2023) 💰 YC Launch Deal: 50% off for 12 months ---------------------------------------- * $20/month per contributing developer (reg. $40) * Use code LAUNCHYC at checkout, **valid until 8/15** ### 🙏 How You Can Help: We're eager for your insights on our product, feature ideas, and security needs - whether through trying our GitHub app, scheduling a meeting, or sharing feedback. * [Install our GitHub app]( * [Schedule a meeting to discuss your application security needs]( * [Share your feedback](mailto:[email protected]) ### 🚀 Get started in 30 seconds: [ 📅 Want to learn more? Schedule a demo: [

Sorcerer

sorcerer

https://sorcerer.earth

San Francisco, CA, USA

Sorcerer builds weather balloons that collect 1000x more data than traditional systems. We have $5.5+ of signed LOIs with the US National Weather Service, the Department of Defense, and the government of El Salvador.

We build weather balloons that collect 1000x more data

Industrials

Industrials -> Climate

S24

Active

Early

https://www.ycombinator.com/companies/sorcerer

https://yc-oss.github.io/api/batches/s24/sorcerer.json

Title: Sorcerer: We build weather balloons that collect 1000x more data | Y Combinator URL Source: Markdown Content: ### We build weather balloons that collect 1000x more data Sorcerer builds weather balloons that collect 1000x more data than traditional systems. We have $5.5+ of signed LOIs with the US National Weather Service, the Department of Defense, and the government of El Salvador. Sorcerer Founded:2024 Team Size:3 Location:San Francisco ### Active Founders ### Austin Tindle, Founder/CEO CEO at Sorcerer. Engineering and product leader of decacorn hardware teams. Previously Head of Product at SumUp, led acquisition of Fivestars (YC W11). ### Maxmillion McLaughlin, Founder/CTO CTO of Sorcerer. Previously founding engineer at aerospace startup Urban Sky where I developed the world’s first stratospheric ballooning systems for ultra high-resolution imaging and real-time wildfire tracking. All things industrial and aerospace hardware for the last decade. ### Alessandro Vecchi, Founder/CPO CPO of Sorcerer. I was employee #3 at Urban Sky, where I built their mission control software ("the best in the world" - US Secretary of Defense) and prediction engine from the ground up. Over 5 years of experience working with/dealing with forecast and weather data. Building software products since I was in high school. ### Company Launches [### Sorcerer - Weather balloons that collect 1000x more data]( Hey YC! We’re [Max]( [Alex]( and [Austin]( the team behind [Sorcerer]( **The problem & background** Hundreds of terabytes of weather data are collected from satellites, ground stations, airplanes, and weather balloons every day. In-situ weather observation has the most impact on weather models, but it’s incredibly expensive to collect. The US National Weather Service spends over a billion dollars annually _just on its network of weather balloons, stations, and aircraft sensors_. Despite this cost, there are still places in the US [where we don't know what the temperature will be two days from now]( And for the 80% of the world that lacks any weather infrastructure? There’s always the [weather rock]( **The solution** We’re launching a global network of persistent weather balloons to provide real-time data in previously unreachable locations. Each balloon remains airborne for over six months, completing ~30 laps around the globe while navigating between sea level and 65,000 feet. We’re able to collect 1000x more data per dollar than current systems, and we use this unique data to train AI weather models that are up to 50% more accurate than today’s best forecasts. [ **The Team** As founding engineer at aerospace startup [Urban Sky]( Max developed the world’s first stratospheric ballooning systems for high-resolution imaging and wildfire tracking. As employee number three at Urban Sky, Alex built their mission control and flight prediction engine from the ground up (“the best in the world” - US Secretary of Defense). Austin was an ultralight flight instructor before he could legally drive and was most recently Head of Product at [SumUp]( shipping highly regulated hardware at scale. **How You Can Help** * **Industry intros:** We’d love to talk with aerospace, commercial aviation, ocean freight, logistics, and commodities trading execs (these folks think weather data is pretty neat) * **If you find one of our balloons near SF**: Sorry! It’s still a work in progress (and please get it back to us) We’d also love to hear from you! Reach out any time at [[email protected]](mailto:[email protected]) Best, Max, Alex, & Austin #### Company Photo ### Hear from the founders #### What is the core problem you are solving? Why is this a big problem? What made you decide to work on it? Hundreds of terabytes of weather data are collected from satellites, ground stations, airplanes and weather balloons every day. A typical weather satellite costs more than $100 million to launch, and the US National Weather Service spends over a billion dollars a year _just on its sparse network of weather balloon stations_. A small handful of organizations then pipe that data into billion-dollar supercomputers to run the world’s most expensive physics simulations four times a day. Despite this cost, there are still places in the US [**where we don't know what the temperature is going to be two days from now**]( And for the rest of the world that lacks adequate weather infrastructure? There’s always the [**weather rock**]( AI models are upending the expensive physics simulations. These models are more accurate, running at higher resolutions for a fraction of the cost. But, like most AIs, weather models are hungry for _data,_ and the reality is that most of the world isn’t covered despite the costly infrastructure.This is because _direct_ observation (as opposed to satellite observation) is both incredibly impactful but also incredibly expensive. This cost leaves **most of the Earth without in-situ observations**. Oceans have especially poor coverage, and no matter how good your model is, you still need to know what’s happening over the Pacific today to predict the weather in SF tomorrow. Sorcerer is deploying a global network of low-cost, persistent weather drones to provide real-time data in previously unreachable locations. Our system collects 30x more observations than the entire global observation infrastructure at 1% the cost. We use this data to train a weather model that’s already competitive with best-in-class traditional weather prediction, while being 10,000x cheaper.

Cheers

cheers-2

https://www.cheersreviews.com

San Francisco, CA, USA

Cheers is building the operating system for in-person businesses, bringing the seamless efficiency of an Apple Store to every local service. We're starting with NFC-enabled employee badges that solve immediate pain points: collecting reviews, accepting payments, attributing customer interactions, and tracking performance. Our customers 5-10x their monthly Google Reviews. - Imagine walking into any local business and experiencing the smoothness of an Apple Store: roaming checkouts, instant transactions, and personalized service - all from a tap on an employee's phone. That's the future we're creating. Our badges are the first step towards a fully integrated, frictionless system for in-person operations. Most tech ignores field workers. We're fixing that. We're digitizing all operations for local services - from home repairs to restaurants. Our vision: tap your phone anywhere to close a sale, leave a review, or verify service, making every interaction as smooth as buying the latest iPhone. - We have over 30 paying customers, including one enterprise pilot. MRR grew from $1,400 to ~$5,000 in two months. We've processed over 3,000 reviews. Founded by Dylan Allen and Ernesto Valencia, CS grads from BYU with experience at Microsoft and Amazon. We're expanding into scheduling, payments, and full workforce management. Our goal: become the central platform for in-person work, serving millions of field employees and revolutionizing how local businesses operate.

The super-app for in-person employees

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/cheers-2

https://yc-oss.github.io/api/batches/s24/cheers-2.json

Title: Cheers: The super-app for in-person employees | Y Combinator URL Source: Markdown Content: ### The super-app for in-person employees Cheers is building the operating system for in-person businesses, bringing the seamless efficiency of an Apple Store to every local service. We're starting with NFC-enabled employee badges that solve immediate pain points: collecting reviews, accepting payments, attributing customer interactions, and tracking performance. Our customers 5-10x their monthly Google Reviews. - Imagine walking into any local business and experiencing the smoothness of an Apple Store: roaming checkouts, instant transactions, and personalized service - all from a tap on an employee's phone. That's the future we're creating. Our badges are the first step towards a fully integrated, frictionless system for in-person operations. Most tech ignores field workers. We're fixing that. We're digitizing all operations for local services - from home repairs to restaurants. Our vision: tap your phone anywhere to close a sale, leave a review, or verify service, making every interaction as smooth as buying the latest iPhone. - We have over 30 paying customers, including one enterprise pilot. MRR grew from $1,400 to $10,000+ in three months. We've processed over 3,000 reviews. We're expanding into scheduling, payments, and full workforce management. Our goal: become the central platform for in-person work, serving millions of field employees and revolutionizing how local businesses operate. Cheers Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Dylan Allen, Founder Dylan Allen is the CEO & Co-Founder of Cheers. Previously, Dylan interned at Microsoft Azure and other Utah-based B2B Tech companies as a product manager & software engineer. Dylan graduated Cum Laude from BYU with a degree in computer science, and originally hails from Dallas, Texas. ### Ernesto Valencia Arevalo, Founder Ernesto is the CTO & Co-Founder of Cheers. Previously, Ernesto interned at Microsoft, Amazon, worked part-time during college at SlingTV among other SWE experiences. He has also conducted research in Artificial Intelligence, Game Theory, and HCI at the Laboratory for Interactive Machines under Dr. Jacob Crandall at his university. Ernesto Valencia Arevalo [Cheers]( ### Company Launches [### Cheers: The super-app for in-person employees]( **TL;DR:** We’re [Cheers]( building the super-app for in-person employees (think home services, restaurants, D2D sales). We’re starting with NFC-enabled employee badges that make key tasks—like collecting reviews and verifying employment—seamless. It’s our first step toward becoming the central platform for in-person work. **🥂 What is Cheers?** ---------------------- _Cheers is on a mission to revolutionize how in-person employees work._ We’re building the super-app that will become the central hub for everything these employees need—from managing schedules and tracking performance to accepting tips and capturing customer feedback—all through NFC technology pre-enabled on smartphones today. **🧨 The Problem We Solve:** ---------------------------- **In-person customer and employee interactions are often inefficient and filled with friction.** Whether it’s collecting feedback, verifying employment, or capturing leads, these processes are typically slow, manual, and prone to errors. This inefficiency not only frustrates employees and customers _but also leads to missed opportunities for businesses to optimize and grow._ Check out how it looks here: [ **Cheers aims to make every in-person interaction seamless, frictionless, and measurable.** We’re starting with NFC-enabled employee badges that go beyond basic identification. These badges enable employees to perform key actions with just a tap—whether it’s gathering a 5-star Google Review, collecting a lead, or verifying employment—right when it matters most, face-to-face with the customer. **🚀 Why We’re Excited About Cheers:** -------------------------------------- We believe that these badges are our wedge into businesses. They address immediate needs (better Google Reviews, etc.) while laying the groundwork for something much bigger. **As employees and owners tap into Cheers’ early features, they quickly see the value in capturing and attributing customer interactions to individual employees.** That means the right employees get the proper credit and feedback, which is actionable at a more granular level. We’ve already seen this open the door to deeper integration across operations. _(**also, Apple** made an announcement about **iOS 18.1** that really excites us, check it out _[_here_]( **Example:** Imagine an employee tapping their badge to allow a customer to leave a 5-star Google Review in just 3 taps: notification, rating, post. Or, picture a customizable employee page where customers can perform multiple actions—like booking services or leaving feedback—all in one place. **📈 Our Traction:** -------------------- We have 25 customers, the majority of which are small businesses, like your local bridal shop or burger joint. We have one enterprise-level client with whom we’ve been running a successful paid pilot for the past 2 months. **We’ve grown our MRR from ~$1,400 on July 1st to ~$4,000 now.** Cheers has already processed over 3,000 internal and external reviews for our customers. Our reporting software is delivering real results—**our average customer has 5-10x’ed their monthly 5-star Google Reviews.** Managers are using it to track performance, run competitions, and provide targeted feedback. **(See the screenshot below for a look at what our customers do when hosting competitions).** **But this is just the start.** We’re building Cheers to be the central operating system for in-person businesses—a platform that will transform how in-person teams operate, from scheduling and payments to customer engagement and beyond. **🙏 Our Ask to You:** ---------------------- We’re looking for intros to folks who manage—or know someone who manages—in-person, customer-facing teams: anything from self-storage to solar sales. Cheers is _especially_ helpful for companies that currently struggle with getting online reviews (Google, TripAdvisor, Thumbtack, etc.). Our first product, the Cheers NFC-enabled employee badge, is already delivering strong results, and with your help, we can build something truly impactful. Request a text from one of us [here]( or schedule a call with our CEO Dylan, [here]( **🤩 The Team:** ---------------- We’re [Dylan Allen]( and [Ernesto Valencia]( – we just moved to SF from Utah to bring Cheers to life. We worked at Microsoft and Amazon, and graduated from BYU where we studied Computer Science. Dylan and Ernesto hail from Dallas, TX and San Salvador, El Salvador, respectively. We’re building something we believe will redefine how in-person businesses are run. Help us make that a reality! And finally, a toast, to people like you. Cheers! 🥂

Saphira AI

saphira-ai

https://saphira.ai

San Francisco, CA, USA

Saphira is the easiest way for hardware products to get safety certified and reach market faster, such as industrial robots and heavy machinery.

Vanta for Hardware

B2B

B2B -> Engineering, Product and Design

S24

Active

Early

https://www.ycombinator.com/companies/saphira-ai

https://yc-oss.github.io/api/batches/s24/saphira-ai.json

Title: Saphira AI: Vanta for Hardware | Y Combinator URL Source: Markdown Content: ### Vanta for Hardware Saphira is the easiest way for hardware products to get safety certified and reach market faster, such as industrial robots and heavy machinery. Saphira AI Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Akshay Chalana, Founder Co-Founder, CEO - Saphira AI - Automating Safety+Systems Engineering: ex-Citadel (real-time trading workflow configuration), Apple (pb-scale dataset management for Siri and Vision Pro), Tesla (made sure batteries didn't explode in the heat of the sun), Meta (PyTorch/ONNX), Hyperloop (blew up and crashed pods) ### Oscar Avatare, Founder Saphira AI (S24) Formerly at Amazon and Twitter. Passionate about helping the next generation of innovative hardware companies build better products faster. Per aspera ad astra. ### Company Launches [### Saphira: Vanta for hardware]( **tl;dr: Hardware companies** use [Saphira]( to easily **generate safety reports** and **get certified to sell** their product. We’re [Oscar]( and [Akshay]( and we’ve built software for hardware and compliance at Twitter, Tesla, Amazon, and Hyperloop. We’re making it really easy to get **UL, CE, ISO, and TÜV** certifications like **ISO 10218** and **ISO 26262** for hardware products. What are safety certifications, and who needs them? --------------------------------------------------- Building **industrial machinery** that can hurt someone, like robot arms and autonomous trucks?  Your **customers will mandate** TÜV safety certification per particular ISO and IEC standards! Building **small appliances** that plug into the grid?  Many **retailers won’t sell your product,** and **commercial buildings won’t install** it unless you’re UL or CE certified! Building a **satellite or communication system**?  The **FCC won’t allow** it to operate without certification! Building **defense equipment** for the government?  **They won’t buy** it unless it’s compliant with the relevant MIL-STDs! Why did we start Saphira? ------------------------- Hardware products can’t get to market because teams **don’t know how to navigate the certification process**. Unlike SaaS, there are no tools like Vanta that tell you exactly what you need to test, change, and re-evaluate in your designs to pass audits. This sucks because every hardware company we talk to gets stuck in **rework cycles** with their consultants and auditors that they call **“horrendous and intimidating”** (real customer quote). We've seen this at Amazon and Tesla, where engineers don't know when a change breaks certification. Each iteration requires them to work with consultants to understand what broke, lengthening the time to launch. We created Saphira to empower these innovators to **rapidly** build complex products **that actually work**, all while achieving and maintaining compliance with the **evolving regulatory landscape**. What does Saphira do? --------------------- Saphira tells you exactly what you need to account for in your **design and testing** to **achieve and maintain** hardware safety certification and compliance. We’re a centralized data source for the **requirements, tests, and standards** that are relevant to your product, making it straightforward to **identify and triage safety and compliance risks** without wasting months of time and thousands of dollars on **slow safety consultants**. Saphira automates **traceability**: regulatory reports by engineers that prove that requirements like “the robot arm won’t shock an attendant if touched while in motion” were implemented and validated. These reports are **painful and tedious** to produce. By maintaining references to all necessary source information, Saphira can automatically generate these reports, and even automatically identify which additional standard clauses are relevant to each requirement or design value. Who is Saphira for? ------------------- Saphira is for engineers at hardware organizations of any scale and size: * The **safety engineering team** at an industrial robotics company can use Saphira’s **compliance report generator** (i.e. ISO-12100 and ISO-10218) to **achieve certification** to operate in the US. * An **engineering leader** at a hardware startup that needs UL certs to sell to more customers can use Saphira to speed up their process by months by avoiding hiring a slow safety consultant. * A **systems engineer** at an aerospace/defense company can use Saphira’s **AI tooling** to understand **how various military and government standards** (such as MIL-STD-810 or ISO-26262) are relevant to each facet of their specs and models. Ask --- * We are experts in certifications for hardware companies.  If you think you might need this but aren't sure how to approach it, we are happy to help for free — [just reach out](mailto:[email protected]). * If you know other engineering leaders seeking to improve their certification processes, [connect us](mailto:[email protected])! We are most interested in talking to leaders in industrial robotics, energy systems, autos, aeros, and defense. * If you worked at a larger hardware company before and saw any of these issues, we would love to talk just to learn from your experience * We would love product feedback: [book a demo]( ### Other Company Launches ### Saphira AI - Orchestrating safe hardware engineering Intelligent engineering design management for safety-critical products [Read Launch ›]( ### YC S24 Application Video

Thyme

thyme

https://www.meetthyme.com/

San Francisco, CA, USA

There are 300k+ financial advisors in the US using a digital Rolodex to manage their client relationships, and it's time for a makeover. At Thyme we are building the OS for financial advisors that allows them to automatically take notes in meetings (in-person, online, and phone calls), manage their client records with the help of AI, and offer a client portal where clients can track their assets, manage tasks, and even pay for their services. Our users have replaced their legacy CRMs and billing solutions with us while deepening their clients' relationships thanks to our client portal and AI tooling that puts rich context at their fingertips.

The OS for financial advisors

B2B

B2B -> Finance and Accounting

S24

Active

Early

https://www.ycombinator.com/companies/thyme

https://yc-oss.github.io/api/batches/s24/thyme.json

Title: Thyme: The OS for financial advisors | Y Combinator URL Source: Markdown Content: ### The OS for financial advisors There are 300k+ financial advisors in the US using a digital Rolodex to manage their client relationships, and it's time for a makeover. At Thyme we are building the OS for financial advisors that allows them to automatically take notes in meetings (in-person, online, and phone calls), manage their client records with the help of AI, and offer a client portal where clients can track their assets, manage tasks, and even pay for their services. Our users have replaced their legacy CRMs and billing solutions with us while deepening their clients' relationships thanks to our client portal and AI tooling that puts rich context at their fingertips. Thyme Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Daniel Lantz, Founder I am originally from Texas and currently reside in Richmond, VA. Prior to co-founding Thyme, I worked as a software engineer at Capital One, trovata.io, and meetcampfire.com (YC S23). In my free time, I enjoy Brazilian Jiu-Jitsu, chess, and spending time with my wife and son. ### Fernando San Martin, Founder Electrical engineer from Spain. I moved to SF to start my career. For the most part, I've worked at the intersection of finance, software, and data at places like Capital One, Zippi (YC s19), and Trovata. ### Company Launches [### 🌿 Thyme: meeting intelligence for financial advisors]( **tl;dr: **[**Thyme**]( enables financial advisors to run 10x better meetings. We cover meeting prep, transcription, summarization, CRM syncing, and document parsing. Please introduce us to any financial advisor in your network!** 🧑‍🤝‍🧑 About us ----------------- We are Fernando San Martin (left), and Daniel Lantz (right). We've worked together at Capital One, [trovata.io]( Campfire (YC S23) and are now Thyme co-founders. Our mission is to build better software tools for financial advisors. **❗ The Problem** ----------------- Financial advisors spend hours every day preparing, running, and following up on client meetings. Their attention is split between engaging with the client and taking meaningful notes, and often, these notes are incomplete or non-existent. After finishing the meeting, the advisor has to expand on the few scattered words that they jotted down to create follow-up tasks and update their CRM with the new client insights. ✅ **Our Solution** ------------------ Thyme solves all of the above. Using our app, the advisor can see their upcoming meetings, for which they can prepare by viewing a holistic profile about the client and even asking natural language questions. Once in the meeting, our app uses state-of-the-art AI technology to transcribe, summarize, generate action items, and draft follow-up emails. Finally, our powerful integrations with the rest of the advisors’ tech stack also allow us to initiate workflows and send tasks and notes to other systems. Any financial documents submitted by the client are parsed and integrated into their profile. Our users describe the solution as a game-changer that saves them hours each week while creating much richer and more accurate notes and tasks that sync seamlessly with their stack. 🙏 **Request** -------------- * If you like what we are building, upvote, comment, and follow us on [Linkedin.]( * If you know of any financial advisors in your network, ask them if they would be interested in chatting. Here is a potential blurb you can use to introduce us: "[Thyme]( is changing how financial advisors run online meetings. Would you like to hear more directly from the team?" * Like and re-share our upcoming public launch with your network later this week. * Reach out to us at [[email protected]](mailto:[email protected]) if you have ideas or suggestions or would like to simply meet. #### YC Sign Photo #### Company Photo

Argil

argil

https://www.argil.ai/

Paris, Île-de-France, France

The video generation engine for the creator economy

Consumer

Consumer

S24

Active

Early

https://www.ycombinator.com/companies/argil

https://yc-oss.github.io/api/batches/s24/argil.json

Title: Argil: The video generation engine for the creator economy | Y Combinator URL Source: Markdown Content: ### The video generation engine for the creator economy Argil Founded:2023 Team Size:8 Location:Paris, France ### Active Founders ### Laodis Menard, Founder Create content for social media with your deepfake on Argil.ai Fascinated by stories, creative (wrote a SF novel and directed a thriller) ### Brivael Le Pogam, Founder Hacker building the best video avatar for content creators. ### Company Launches [### Argil - Generate videos with your AI clone in 2 minutes 🤳]( **tl;dr:** You’re probably a great dev but a terrible marketer, and you should use video because we’re making it cheap and easy thanks to our SOTA deepfake model and platform. [Check out my video message to you:]( ### **The problem** **Video is expensive** to produce. It is used for branding rather than performance and is usually the last resort employed after everything is exhausted (ads, SEO, etc.). Yet, it is the **highest performing media** out there. It can engage leads and clients (branding, masterclasses), sell (UGC), and make internal documents more engaging (FAQ, trainings). ### **The solution** By building a **SOTA deepfake model** and a **platform to generate videos**, we’re making it insanely easy to create and grow your audience with video, regardless of your goals: * **Clone yourself** with 2 minutes of video, or use one of our avatars * Input a text or voice script * Generate a **fully edited, engaging video** complete with subtitles & media in a few clicks All of it is also **available via our API** 👀 **What’s magic?** * Make it engaging with body language control & camera angles * Preview your video before generating it * Get a ready-to-post video in 2 minutes (subtitles, media, etc.) ### **Use cases** **Branding** Social media is where you can do the most business. **Even if you’re shy**, you should leverage video on social media to share stories, knowledge, and engage. **If you record ONE good 2-minute video** of yourself, you’re **set for life.** We just released a feature that will help you generate your **first LinkedIn video** and get ahead of the release of their “shorts” feature - try it out! **Education** One of the most effective marketing strategies is to **educate your prospects** on how to solve their pain - you can use lead magnets (free training videos in exchange for their email or action on their part), or generate video documentation on your platform to convert/upsell. **Promotion (screencasts, UGC)** Having influencers sell your product is a huge pain: you need to find them, brainstorm ideas with them, pay them, give them an affiliate plan, etc. How cool would it be to generate **100 ads for your product effortlessly**? That’s what we allow doing with our UGC avatars. **Entertainment** YouTuber Kwebblekop (15m subscribers) has created an automated pre-production pipe that then calls our API. He generates **1000 shorts/month** with the platform, and his channel is growing with an entirely automated workflow. ### **The founders** We met 10 years ago while working at French scale-up Teads. After our first failed startup together in 2016, I (Laodis) worked on several projects involving video: movies, masterclasses, TikTok (25k subscribers)… But during all of it, **the hardest part was production**. Working on AI projects last year, it became obvious that the single pain that was the most meaningful to solve was helping people access video without the production costs and effort getting in the way. ### **Our Asks 👇** * You can generate your first videos right now by [filling out this form]( to **get an onboarding code** and pinging me on Twitter ([@LaodisOfficial]( * If you want to play around **with our API**, reach out to [@BrivaelLp]( on Twitter

Dimely

dimely

https://dimely.co/

San Francisco, CA, USA

Dimely automates manual billing workflows, saving billing teams hundreds of hours every month We do this by doing the data work, including converting contracts to invoices, calculating overages, reconciling payments, and streamlining revenue recognition We pride ourselves in our fast time to value. Dimely seamlessly slots into your existing workflows without requiring a rip and replace, meaning you'll immediately see results from Day 1. Stop spending hours on billing today!

AI Copilot for B2B SaaS Billing Teams

B2B

B2B -> Finance and Accounting

S24

Active

Early

https://www.ycombinator.com/companies/dimely

https://yc-oss.github.io/api/batches/s24/dimely.json

Title: Dimely: AI agents for B2B SaaS Finance Teams | Y Combinator URL Source: Markdown Content: ### AI agents for B2B SaaS Finance Teams Dimely builds AI agents for B2B SaaS finance teams. Our agents automate time-consuming finance workflows so teams can stay lean and realize revenue faster. We do this by doing the data work, including converting contracts to invoices, calculating overages, reconciling payments, and streamlining revenue recognition We pride ourselves in our fast time to value. Dimely seamlessly slots into your existing workflows without requiring a rip and replace, meaning you'll immediately see results from Day 1. Stop spending hours on billing today! Dimely Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### David Pang, Founder/CEO David is co-founder and CEO of Dimely. Before Dimely, David led Business Operations at OneSignal (YCS11) where he was heavily involved with billing. He also co-founded (with Peter) SheetWhiz, a productivity tool for Sheets used by thousands of finance startup professionals every month. He began his career as a tech private equity investor at Silver Lake. He graduated summa cum laude from the Wharton School at the University of Pennsylvania where he studied finance and accounting ### Peter Sun, Founder/CTO Peter is the co-founder and CTO of Dimely. Before Dimely, Peter was the founding Data Scientist at OneSignal (YCS11) where he built data infrastructure for billing. He co-founded (with David) SheetWhiz, a Chrome Extension that brings Excel features to Google Sheets used by thousands of finance professionals. Peter started his career at Meta where he built onboarding experiences for SMB advertisers. Peter graduated from the University of Waterloo with a degree in Electrical Engineering. ### Company Launches [### Dimely 🪙 - AI copilot for B2B SaaS Billing teams]( **TL;DR: **[**Dimely**]( automates manual billing workflows, saving billing teams hundreds of hours a month. Dimely integrates with your existing tech stack, meaning we don't require a rip and replace, and you'll immediately see results from Day 1.** 👋 Hi all - we’re [David]( and [Peter,]( and we’re on a mission to give billing teams their time back. After spending far too many hours doing billing ourselves in our previous roles, we built Dimely as the copilot we wish we had, doing the manual work while keeping a human in the loop to ensure 100% accuracy. **❌ Problem: Billing for B2B SaaS is manual** --------------------------------------------- This is because: * **Contracts are complex.** Contracts have multiple SKUs, payment schedules, pricing models, and billing metrics * **Contracts are unstructured.** Terms are often bespoke and written in by sales teams trying to close deals * **Data lives in different places.** Usage-based pricing models in particular, require matching contract data in your CRM with usage data from your data warehouse that is not always accurate or structured in a way that works well for billing * **Finance needs are deprioritized.** Engineering typically prioritizes product building and deprioritizes finance initiatives * **High finance tech debt.** Billing systems that work early on remain in place for too long because it takes 4+ months to migrate to another system, resulting in a high number of manual workarounds, typically in spreadsheets **🏁 Solution: AI copilot for billing teams easily implemented within existing workflows (no engineers required)** ------------------------------------------------------------------------------------------------------------------ This is where [Dimely]( comes in. We take the tasks that take the most time and automate them while keeping you in the loop. Dimely’s primary areas of focus are: **✅** Converting contracts to invoices **✅** Calculating overages and invoicing customers (via your billing solution) **✅** Payment reconciliations **✅** Setting up proper revenue recognition What makes Dimely unique: * **High time to value.** We integrate within your existing workflows - no need to rip and replace anything * **No engineers required.** Set up Dimely through our UI and with natural language * **Flexible.** We know billing is heterogeneous, and various companies do things differently. Dimely’s AI copilot understands its workflows and won’t break if something unusual comes up **🎇 How it Works:** -------------------- Here’s a short demo of a fairly simple but time-consuming workflow where Dimely can assist: creating invoices in your billing solution based on signed contracts [ **📢 Our Ask:** --------------- * Let us know if you know anyone spending significant time with billing. Specifically, Controllers, Revenue Accountants, and Heads of Finance at B2B SaaS startups. We’d love to [chat]( * Follow us on [LinkedIn]( - we have some pretty good content! #### Company Photo

Glasskube

glasskube

https://glasskube.dev/

Vienna, Vienna, Austria

Glasskube is building a new packaging ecosystem for Kubernetes. With our Open Source package manager, we aim to replace Helm in the long run and give users a Docker-like experience to build, extend, deploy and distribute Kubernetes applications. We're also working on a cloud-native templating language so anyone building packages for Kubernetes can use a DSL and won't have to use the Go templating engine anymore.

The next generation Package Manager for Kubernetes

B2B

B2B -> Infrastructure

S24

Active

Early

https://www.ycombinator.com/companies/glasskube

https://yc-oss.github.io/api/batches/s24/glasskube.json

Title: Glasskube: Open Source Package Manager for Kubernetes | Y Combinator URL Source: Markdown Content: ### Open Source Package Manager for Kubernetes Glasskube is an Open Source Package Manager for Kubernetes. Kubernetes is highly fragmented, making it nearly impossible to deploy applications across multiple environments with current tooling. We help software companies sell to enterprises and deploy to private infrastructure, different cloud environments, and on-premises. The founders have been best friends for over 15 years, lived together in college, and even had a startup together. Philip was the CTO of a European Kubernetes-based company that was acquired in 2022. Glasskube Founded:2023 Team Size:5 Location:Vienna, Austria ### Active Founders ### Louis Weston, Founder/President Business, Law, and Compliance background. Before Glasskube, I advanced from intern to legal counsel at a fintech startup with over 100 employees in three years, negotiating contracts with Apple, Deutsche Bank, and Volkswagen, and building a compliance and security system according to ISO standards. ### Philip Miglinci, Founder/CEO Open-Source and Kubernetes ❤️ Ironman finisher 🏊 🚴 🏃 Before working on Glasskube, I built: - Mission-critical alerting infrastructure in Europe as CTO - PCR testing and analytics software as part of my Master thesis ### Company Launches [### Glasskube - Open source package manager for Kubernetes]( Hi everyone! We’re [Philip]( and [Louis]( the co-founders of [Glasskube]( **TLDR:** Glasskube is an open source package manager for Kubernetes that makes deploying, updating, and configuring packages on Kubernetes **20x faster than** tools like **Helm or Kustomize.** Inspired by the simplicity of Homebrew and npm, our goal is to build enterprise-ready infrastructure software that can run mission-critical workloads across the more than 3 million Kubernetes clusters globally. **Glasskube is fully open source.** Check out our repository on [GitHub]( (We'd appreciate a star 🌟) **The Backstory & Problem** --------------------------- Kubernetes packages have been growing rapidly, from around 100 packages in 2016 to over 800 packages on the CNCF landscape today. With more than 7 million developers working with Kubernetes, Helm, an open-source tool created during a hackathon nine years ago, has become the go-to solution. However, Helm struggles to keep up with the increasing demand, leading to complex workflows and non-standardized solutions. Over the past five years, writing and deploying packages on Kubernetes, we found ourselves defaulting to Helm despite its shortcomings due to a lack of alternatives. While working on other cloud-native projects, users consistently highlighted common pain points with Helm. This realization prompted us to tackle the larger issue of package management in Kubernetes, leading to the development of Glasskube. **What is Glasskube?** ---------------------- Glasskube brings the experience of Homebrew or npm to your Kubernetes cluster. It simplifies installation, updates, and configuration, making it easier and faster to manage your packages. **Focusing on simplicity and reliability with our CLI and UI** Easily install packages in your cluster via the Glasskube UI, where all packages are conveniently located, eliminating the need to search for a Helm repository. **Safe Package Updates** Preview and perform pending updates to your desired version with a single click (or CLI command). All updates are pre-tested by the Glasskube test suite. **Package configurations** Configure packages with typesafe input values via the UI or interactive CLI questionnaire. Inject values from other packages, ConfigMaps, and Secrets easily. No more untyped and undocumented values.yaml files. **GitOps Integration** Interested? Check out our [Demo.]( **Our Vision** -------------- Currently, we are focused on enhancing the user experience, aiming to save engineers as much time as possible by building on top of existing technologies. In the future, we plan to develop an entirely new packaging and bundling format for all cloud-native packages. This will provide package developers with a straightforward way to define how to install and configure packages, offer simple upgrade paths, and enable us to provide feedback, crash reports, and analytics to every developer working on Kubernetes packages, similar to how Android supports Play Store developers. **Presignup - Glasskube Cloud** ------------------------------- We are currently working on Glasskube Cloud, for which you can now [pre-signup]( Key features include: * Cloud Hosted Private Packages * Security Alerts and Update Notifications * Advanced GitOps Integrations (intelligent insights on upgrade success probability, smart config migration suggestions, enhanced open PRs showing detailed changes at manifest/resource level) **Our Asks** ------------ We love talking to engineers and anyone interested in DevOps who uses Kubernetes. You can schedule a call with Louis and Philip [here]( If you find this interesting and would like to try it out: * Presignup for [Glasskube Cloud]( * Check out our repository [on GitHub]( #### Company Photo

Theseus

theseus

https://theseus.us

We've built a Visual Navigation System (VNS) that can be retrofit onto any drone that mimics a GPS. Our VNS uses cameras and an accelerometer/gyroscope as well as reference satellite imagery for positioning. The system is entirely self contained, can't be jammed and has no RF signature.

GPS Denied Navigation for Drones

Industrials

Industrials -> Drones

S24

Active

Early

https://www.ycombinator.com/companies/theseus

https://yc-oss.github.io/api/batches/s24/theseus.json

Title: Theseus: GPS Denied Navigation for Drones | Y Combinator URL Source: Markdown Content: ### GPS Denied Navigation for Drones We've built a Visual Navigation System (VNS) that can be retrofit onto any drone that mimics a GPS. Our VNS uses cameras and an accelerometer/gyroscope as well as reference satellite imagery for positioning. The system is entirely self contained, can't be jammed and has no RF signature. Theseus Founded:2024 Team Size:3 Location: ### Active Founders ### Ian Laffey, Founder GPS-denied nav for drones ### Sacha Levy, Founder UAV nav in gps denied environments. ### Carl Schoeller, Founder Building Theseus – defense systems independent of GPS. Before that designed and built around 2000 unique parts. Some of those are in mass production on Tesla Cybertruck and some in an Apple product. ### Company Launches [### Theseus: GPS Denied Navigation for Drones]( Six months ago we did a hackathon, this month we sold to SOCOM. Our Visual Navigation System (VNS) uses cameras, an inertial measurement unit, and pre-loaded satellite imagery to determine it’s position and send a fake GPS signal to any drone. Integrations typically take <30min. ### How It Started In February, [Carl]( was a Senior at Stanford, [Sacha]( was a PhD student at Yale, and [Ian]( was working with a British celebrity. After our post went viral, we spent weeks talking with troops on the ground, drone manufacturers, and DoD. We learned that: \- The front lines of Ukraine are 100% GPS-denied \- Cheap, DIY drones are extremely effective but typically rely on GPS \- Existing GPS alternatives are extremely heavy and extremely expensive We set out to create a low-cost, low-weight replacement for GPS that works with any drone. ### How It’s Going Our first customer in the Special Forces has enabled us to push our VNS to its limits, and the results have been impressive. Here are a few excerpts from a memorandum written by the Captain we work with: _“Having tested the Theseus’ VizNav system for a week, I am convinced that their technology presents the best solution for a current, and ever-expanding urgent operational need.”_ _“As a company, Theseus has shown a willingness to seek the ground truth from the operator and make improvements based on these recommendations immediately. The progress they have made with limited funding gives me incredible confidence that they act quicker and more efficiently than any other vendors we have worked with.”_ Due to the sensitive nature of our product, technical details and customer-specific information in this post will be limited. We are happy to discuss results in more detail privately. ### Going Forward America is way behind. The current demand for drones in Ukraine is 1-2 million yearly. Every country in the world sees this and is in the process of building out its arsenal. Our product affordably enables an entirely new capability set for every one of these platforms. We have sold units to multiple customers, have six-figure LOIs from large drone manufacturers, integrations scheduled, and eight figures in funding for our VNS requested by the DoD for FY25. The West needs to build millions of drones, and all of those drones need Theseus.

VideoGen

videogen

https://videogen.io

San Francisco, CA, USA

VideoGen makes it easy for anyone to create professional, copyright-free videos in seconds. Today, we serve thousands of customers, from individuals and small businesses to teams at Google and ByteDance. Making videos is hard, time-consuming, and expensive if you hire someone to do it for you. With VideoGen, it only takes a few clicks!

Generate professional videos in seconds

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/videogen

https://yc-oss.github.io/api/batches/s24/videogen.json

Title: VideoGen: AI that can create and edit professional videos | Y Combinator URL Source: Markdown Content: ### AI that can create and edit professional videos VideoGen makes it easy for anyone to create professional, copyright-free videos in seconds. Today, we serve thousands of customers, from individuals and small businesses to teams at Google and ByteDance. Making videos is hard, time-consuming, and expensive if you hire someone to do it for you. With VideoGen, it only takes a few clicks! VideoGen Founded:2023 Team Size:3 Location:San Francisco ### Active Founders ### Anton Koenig, Founder Co-founder and CEO of VideoGen. I previously interned at AWS and dropped out of UMass Amherst Computer Science to focus on VideoGen full-time. ### David Grossman, Founder I'm the cofounder and CTO of VideoGen (S24), a platform for generating professional videos using AI. Before this, I studied Computer Science and Applied Math at Brown and interned at Palantir. ### Company Launches [### VideoGen – Generate professional videos in seconds]( [ **TLDR:** [VideoGen]( makes it easy for anyone to create professional, copyright-free videos in seconds. Today, we serve thousands of customers, from individuals and small businesses to teams at Google and ByteDance. Making videos is hard, time-consuming, and expensive when outsourced. With VideoGen, it only takes a few clicks! 👋 Hello, we’re Anton and David. We bootstrapped VideoGen out of our college dorm rooms to over 1 million users in under a year. We met at a tennis academy in middle school and became friends through our shared passion for building web apps and playing tennis. Now, we’re changing how the world creates videos with the help of AI. ### **The Problem: Creating videos is tedious, painful, and hard.** Making videos is hard. Each step of the process—scripting, recording voice-overs, finding and licensing high-quality footage, and editing—is extremely time-consuming. Video editing products today are hard to use and have too many buttons. Paying someone to make videos for you is expensive and slow. ### **The Solution: AI that creates videos for you.** VideoGen can automatically edit professional videos together in a few seconds, saving our customers time and the pain of hiring a video editor. In just a few clicks, VideoGen can: 🧠 Take in company/brand knowledge and media ✍️ Write a video script 🗣️ Generate a lifelike voice-over ✂️ Edit together relevant, copyright-free footage 💬 Add captions VideoGen does this all in over 100 languages and accents for a global user base. Once your video is done, you can add your finishing touches in an easy-to-use, web-based video editor. ### **Try the API – Get $1,000 in free credits (YC Deal)** Are you interested in integrating VideoGen with your app? Reach out to us at [[email protected]](mailto:[email protected]) #### YC Sign Photo ### Hear from the founders #### How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?) 👋 Hello, we’re Anton and David. We bootstrapped VideoGen out of our college dorm rooms to over 1 million users in under a year. We met at a Tennis Academy in middle school and became friends over a shared passion for building web apps and playing tennis. Now, we’re changing the way the world creates videos with the help of AI. #### What is the core problem you are solving? Why is this a big problem? What made you decide to work on it? Making videos is hard. Each step of the process—scripting, recording voice-overs, finding and licensing high-quality footage, and editing—is extremely time-consuming. Video editing products today are hard to use and have too many buttons. Paying someone to make videos for you is expensive and slow.

Midship

midship

https://midship.ai/

San Francisco, CA, USA

We extract and organize structured data from documents, such as PDFs, excel sheets for non-engineers. A CRE firm uses us to pull data from several deals to populate their underwriting spreadsheets in 5 minutes, instead of hours.

Extract, search and analyze data in your documents

B2B

B2B -> Finance and Accounting

S24

Active

Early

https://www.ycombinator.com/companies/midship

https://yc-oss.github.io/api/batches/s24/midship.json

Title: Midship: Extract, search and analyze data in your documents | Y Combinator URL Source: Markdown Content: ### Extract, search and analyze data in your documents We extract and organize structured data from documents, such as PDFs, excel sheets for non-engineers. A CRE firm uses us to pull data from several deals to populate their underwriting spreadsheets in 5 minutes, instead of hours. Midship Founded:2024 Team Size:3 Location:San Francisco ### Active Founders ### Kieran Taylor, Co-Founder Co-founder of Midship - Extract, search and analyze data in your documents. Spent 7 years building data platforms at Instacart, Amazon & Deloitte. Experienced every pain point possible working with data-focused teams. From the UK, avid traveller, mediocre tennis player, competitive vimmer. ### Aahel Iyer, Founder Former indie rock guitarist turned product designer. Shaped user experiences at Lyft, Faire, and Dashworks, specializing in financial products, and gen AI. Moved from India to the US at 18. Now using diverse insights to build user centric products. ### Max Maio, Co-Founder Shipped software loaning billions at PayPal; built an AI-powered app from zero to one at Dashworks. Snowboarder, surfer, and espresso machine fixer. ### Company Launches [### Midship - Extract documents straight to your spreadsheets]( Hey YC! We're Aahel, Kieran, and Max from [Midship]( We became friends while working together at our previous company. We grew up in different corners of the world, but quickly found common ground over intense FIFA matches, Blue Bottle Coffee and love for building impactful software. We're so thrilled to be on this journey. ### **TLDR** [Midship]( extracts information from complex documents and inputs it directly into your preferred spreadsheets, significantly reducing manual data entry. ### **The Problem** Enterprise professionals waste hours manually extracting data from documents into spreadsheets. Whether it's loan applications or legal contracts, this process steals hours from valuable analysis. ### **Our Solution** 1. Upload your source documents (PDFs, reports, etc.) 2. Upload your target spreadsheet template 3. Midship extracts the required data and maps it to your template 4. Download your fully populated spreadsheet, ready for analysis in minutes ### **See it in action** [ ### **Our Ask** Know senior finance professionals in these areas? * Hedge Funds or Asset Management * Investment Banking * Private Equity We've made it easy to help! Just drop their names, LinkedIn URLs, or emails in our [quick form]( Your referrals are incredibly valuable to us as we grow. ### **We are launching a Startup Tier** We're excited to introduce our Startup Tier, designed to support early-stage companies. While we're starting small to ensure quality support, we'd love to hear from you if you're moving data from PDFs to Excel/GSheets or structured JSONs. [

Ares Industries

ares-industries

https://aresindustries.com

El Segundo, CA, USA

Building low-cost cruise missiles

Industrials

Industrials -> Aviation and Space

S24

Active

Early

https://www.ycombinator.com/companies/ares-industries

https://yc-oss.github.io/api/batches/s24/ares-industries.json

Title: Ares Industries: Building low-cost cruise missiles | Y Combinator URL Source: Markdown Content: Ares Industries: Building low-cost cruise missiles | Y Combinator =============== []( "Y Combinator") [About]( [What Happens at YC?]( Interview Guide]( Blog]( [Companies]( [Startup Directory]( Companies]( Directory]( YC]( [Startup Jobs]( [All Jobs]( Engineering]( Operations]( Marketing]( Sales]( Internship Program 2024]( Job Guide]( Coaching]( Startup Jobs Blog]( [Find a Co-Founder]( [Library]( [SAFE]( [Resources]( [Startup School]( Investors]( News]( Open main menu Apply for **W2025** batch.[Apply]( "Apply for W2025 batch.") [Home]( Industries Ares Industries =============== Building low-cost cruise missiles [S24]( Active [aerospace]( * * * [Company]( [Jobs](  [ * * * ### Building low-cost cruise missiles Ares Industries Founded:2024 Team Size:2 Location:El Segundo, CA Group Partner:[Jared Friedman](  []( "LinkedIn profile") []( "Twitter account") ### Active Founders ### Alex Tseng, CEO and Founder Missiles are cool Alex Tseng [Ares Industries](  []( "LinkedIn profile") ### Devan Plantamura, Founder I am a serial entrepreneur in the defense technology sector with a passion for safety, ethics, and compliance no matter the product. I have experience in operational excellence, business development, compliance, and deep understanding of AI, NLP, and Hardware. Ares Industries is focused on building the next generation of smart munitions. Devan Plantamura [Ares Industries](  []( "LinkedIn profile") ### Company Launches [### Ares Industries - Building low-cost cruise missiles 🚀]( Hello, we’re Alex and Devan, and we’re building the low-cost cruise missiles the United States needs! **What we’re building** Ares is building a new class of anti-ship cruise missiles. We are going to deliver the capabilities that the DoD wants in a form factor that’s 10x smaller and 10x cheaper. Our missiles will be compatible with existing launch platforms, deliver smaller payloads at high subsonic speeds, and take out ships hundreds of miles away. We’re focusing on ground and ship-launched variants to start but will expand into air-launched versions, as well as versions with extended ranges and different payloads. **Why we’re doing this** A war with China in the Taiwan Strait would look nothing like what we've seen in Ukraine or the Middle East. Wargames within DoD and military experts are in agreement that the most useful weapons in such a conflict would be long range anti-ship weapons or cruise missiles. The United States is not adequately prepared. In a potential conflict, our stockpiles will run out in weeks, and we currently don’t have the industrial capacity to build at a rate that could win a war, much less deter China. Higher-ups in DoD are waking up to the problem, and generals such as Major General Cameron Holt and officials such as Chairman of the House Select Committee on China Mike Gallagher have echoed this sentiment. Recent conflicts in the Middle East and Ukraine have shown that our weapons are too large, too expensive for the wars of today. Existing anti-ship cruise missiles are low-volume, multi-million dollar, 3,000 lb behemoths meant to take out large cruisers and destroyers. But Chinese shipbuilding capability is far superior to ours, and we can’t make enough anti-ship missiles to counter. In addition, huge missiles aren’t needed to take out the smaller corvettes and frigates that make up the People’s Liberation Army Navy. And $3M missiles make even less sense to take out swarms of $200k small, unmanned surface vessels. The DoD needs smaller, less expensive cruise missiles, and lots of them! But no one has delivered on that promise yet. **How we know how to do this** Both of us have worked at multiple defense and munitions startups. [Devan]( is an Army Psychological Operations and Navy combat veteran and has deep experience in operations and business development at well-funded defense startups. [Alex]( was the first engineering hire at an earlier missile startup, built autonomous systems at Rivian and Stanford, and worked on turbojets at CSULB. **Where we’re at now** We spent the summer building multiple prototypes and flight-testing them in the Mojave. In 11 weeks, we went from starting the company to flight testing with our own design. We are on track to deliver early working missile systems to our first customers by mid-2025. We believe our vision for anti-ship cruise missiles can help fill the munitions gap that the US faces. At Ares, we would never wish for a war to occur. But by helping to rebuild our defense industrial base, we can make sure the United States is prepared to stop a conflict and save countless lives. **Our Ask** We are looking for extremely talented aerospace engineers to help bring our prototypes to fruition. If you know any aerodynamics, propulsion, or GNC engineers who are interested in building something incredibly cool and want to make a positive difference in geopolitics, please send them our way! Footer ------ [Y Combinator]( "Y Combinator") ### Programs * [YC Program]( * [Startup School]( * [Work at a Startup]( * [Co-Founder Matching]( ### Company * [YC Blog]( * [Contact]( * [Press]( * [People]( * [Careers]( * [Privacy Policy]( * [Notice at Collection]( * [Security]( * [Terms of Use]( ### Resources * [Startup Directory]( * [Startup Library]( * [Investors]( * [SAFE]( * [Hacker News]( * [Launch YC]( * [YC Deals]( ### Make something people want. [Apply]( [Twitter]( © 2024 Y Combinator

Remo

remo

http://getremo.ai

San Francisco, CA, USA

Remo is building AI copilots for onboarding and compliance teams at financial institutions. We automate highly manual processes like KYB and transaction monitoring.

AI for compliance teams at banks

Fintech

Fintech

S24

Active

Early

https://www.ycombinator.com/companies/remo

https://yc-oss.github.io/api/batches/s24/remo.json

Title: Remo: AI for compliance teams at banks | Y Combinator URL Source: Markdown Content: ### AI for compliance teams at banks Remo is building AI copilots for onboarding and compliance teams at financial institutions. We automate highly manual processes like KYB and transaction monitoring. Remo Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Rabii Malik, Founder Former McKinsey consultant and international expansion lead at Telda - now bring AI to financial institutions. ### Hashim Rehman, Founder Co-Founder & CTO at Remo. I previously lead and scaled Pricing Engineering at Careem, built multiple consumer products at Deliveroo and consulted as a full-stack engineer for British Airways. ### Company Launches [### Remo: AI copilot for onboarding and compliance teams]( ### **TL;DR** * Onboarding and compliance teams at financial institutions struggle with scale, cost, and accuracy. * We’re building modular, highly auditable copilots to empower these teams to get through cases faster, reduce cost, and allow them to focus on more complex cases. ### **👨‍💻Team** Hi, we’re Rabii and Hashim, and we’re building [Remo]( We’ve seen firsthand how financial institutions continue to struggle with manual workflows as they onboard customers and fight financial fraud— we’re here to help! CEO, [Rabii]( has a Philosophy, Politics, and Economics Degree from the University of Oxford. He was previously a consultant at McKinsey and Company. Over the last two years, he has led international expansion at Telda, a neobank in the Middle East, helping them scale from 0 to 500,000+ customers. CTO, [Hashim]( has a Computer Science degree from King’s College London. He’s spent the last few years leading and scaling Pricing Engineering at Careem for millions of customers. Before that he built multiple consumer products at Deliveroo and consulted as a full-stack engineer for British Airways. ### 🆘 **Problem** Large financial institutions spend millions of dollars a year on manual processes in their onboarding, AML, and compliance teams, facing a number of challenges as they scale: 💰 **Cost:** Costs associated with onboarding and compliance teams can quickly spiral out of control as financial institutions grow. ❗ **Quality:** Highly manual processes without automation create scope for increased human error. ⌛ **Time:** Fast-growing financial institutions are faced with customer backlogs and struggle to meet their SLA times. 🙅‍♂️ **Focus:** When financial institutions are building their own tools to support analysts, they have to invest time and resources distracting them from their core product. With increased regulatory enforcement and the rise of AI-enabled fraud, costs continue to grow rapidly. In the U.S. alone, $61 Billion is spent per year on financial crime compliance across staff and software - globally, the number is over $300 Billion of annual spend! ### **🛠️** **Solution** A full suite of modular, highly auditable copilots to empower the whole onboarding and compliance organization. **KYC/KYB Onboarding Copilot**: Automatically triages flagged alerts, conducts independent research, generates requests for information, and prepares documentation for human review to ensure that customers are onboarded faster and more accurately. **AML Copilot:** Auto-completes suspicious activity reports, conducts ongoing due diligence, and facilitates internal reporting of AML cases, empowering teams to focus on more complex cases. **Monitoring Copilot:** Automates internal information requests, aggregates insights from sampling efforts and generates regular reports for management teams. ### **✨ ASK** If you or someone in your network oversees compliance operations at a financial institution, we’d love to chat with you! You can book an introductory call [here]( or email [[email protected]](mailto:[email protected]).

Keet

keet

https://trykeet.com

Keet provides APIs for any website. Developers use us to connect their customers accounts and perform actions on their behalf. Whether it be legacy software or consumer sites, Keet manages your users authenticated sessions and provides stable integrations on any web platform. Now your AI Agent can access healthcare data or send automated messages on LinkedIn via Keet’s API or SDK.

APIs for any website

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/keet

https://yc-oss.github.io/api/batches/s24/keet.json

Title: Keet: APIs for any website | Y Combinator URL Source: Markdown Content: ### APIs for any website Keet provides APIs for any website. Developers use us to connect their customers accounts and perform actions on their behalf. Whether it be legacy software or consumer sites, Keet manages your users authenticated sessions and provides stable integrations on any web platform. Now your AI Agent can access healthcare data or send automated messages on LinkedIn via Keet’s API or SDK. Keet Founded:2024 Team Size:2 Location: ### Active Founders ### Zack Ashen, Founder Co-Founder @ Keet, Cornell CS. ### Tommy Tsai, Founder Co-Founder @ Keet, Cornell CS. ### Company Launches [### Keet - APIs for any website]( **TLDR:** [Keet]( provides APIs for any website. Developers use us to connect their customers’ accounts and perform actions on their behalf. **Why does this exist?** ------------------------ It is difficult to build automations with web platforms, especially those that require authentication. We make seamless and stable integrations your competitive advantage. **Who is this for?** -------------------- You want to be able to take actions on your users’ behalf. Imagine if you could easily… * Call an endpoint to submit a form instead of writing a Playwright automation * Book Ubers based on your user’s calendar and location * Track activity on properties for future homeowners on Redfin Without… * Storing your end user’s credentials * Dealing with MFA * Writing playwright automations that break when a page gets updated **There are some other solutions out there. How are we different…** 1. No Chrome extension is necessary for your users to authenticate 2. We maintain your users’ authenticated sessions 3. Many of our API abstractions don’t rely on playwright or browser automations **How does it work** -------------------- 1. Link your end user’s account using a Plaid style link component (Web + iOS supported) 2. Call our API endpoints for their connected platforms: ``` keetClient.integrations.venmo.getTransactions() ``` **Our Asks** ------------ If you have any custom integrations you want built, reach out!

Praxos

praxos

https://www.praxos.ai

Miami, FL, USA; Remote

Praxos allows insurance professionals to automate their operations—from filling out applications to crafting proposals and comparing coverages—in one place. This enables risk advisors to stand out and close more deals in an industry where service speed is critical, but mistakes are costly.

AI workbench for insurance teams

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/praxos

https://yc-oss.github.io/api/batches/s24/praxos.json

Title: Praxos: AI workbench for insurance brokers | Y Combinator URL Source: Markdown Content: ### AI workbench for insurance brokers Praxos allows insurance professionals to automate their operations—from filling out applications to crafting proposals and comparing coverages—in one place. This enables risk advisors to stand out and close more deals in an industry where service speed is critical, but mistakes are costly. Praxos Founded:2023 Team Size:3 Location:Miami, FL ### Active Founders ### Lucas Urbisaia, Founder Harvard grad, Argentine-born mathematician. CEO of Praxos, an AI workbench for insurance professionals. Formerly managed risk at Santander's Sales & Trading desk. ### Soheil Sadabadi, Founder Computer Scientist and Al-Khwarazmi wannabe. ### Company Launches [### Praxos – AI workbench for insurance brokers]( ### **TL;DR** [Praxos]( allows insurance brokers to automate their operations—from filling out applications to crafting proposals and comparing coverages—in one place. This enables brokers to stand out and close more deals in an industry where service speed is critical, but mistakes are costly. ### **Problem** Insurance brokers spend 95% of their time reading through hundreds of pages of client files and entering information from one document to another. They must also keep logs of their activities and client information to comply with regulations. Solving these problems requires a system that can extract data from all insurance documents. No solution has ever been able to achieve this, until now. ### **Solution** An AI-powered workspace for insurance workflows that enables brokers to generate any document they need in 3 clicks. Today, Praxos does two things. First, it fills out insurance applications by extracting document data from brokers' clients. [ Second, Praxos condenses the detailed quotes returned by insurance companies into easy-to-read comparison tables that brokers and their clients can use to pick the best policy. For a typical business, these two tasks take more than 20 hours of painstaking work from trained experts. Meanwhile, every change to the data is logged in the background, ensuring that brokers always have a paper trail ready for their auditors. ### **So what?** Insurance professionals using Praxos report: * 90% decrease in time spent generating documents. * Zero time spent logging activity. * Faster deal closure, with service times dropping from weeks to hours. ### **Team** We are a team of AI nerds, ex-risk managers, and product builders who got tired of drowning in insurance paperwork. Instead, we'd rather use technology to help our peers focus on more engaging parts of the job… and make sure that businesses and people have the correct protection to handle catastrophe when it happens. [Lucas]( * Modeled and managed risk at Santander's Sales & Trading desk. * Graduated with a degree in Applied Mathematics from Harvard. [Soheil]( * Modeled Litigation and Cyber risk at Aon, the world's second-largest insurance broker. * Graduated with a degree in Computer Science from Harvard. ### **Ask** We're live and onboarding insurance brokers daily. Know any brokers? We'd love an introduction! Drop us a note at [[email protected].](mailto:[email protected]) Many thanks for your support! ❤️ Lucas & Soheil #### YC Sign Photo

Terra

terra-2

https://useterra.com/

New York, NY, USA

Enter Terra, a new world of product building. Connect with suppliers, expedite sampling, and track production in real-time, removing the uncertainties of overseas manufacturing.

Terra simplifies product design, manufacturing, and sourcing for SMBs

B2B

B2B -> Supply Chain and Logistics

S24

Active

Early

https://www.ycombinator.com/companies/terra-2

https://yc-oss.github.io/api/batches/s24/terra-2.json

Title: Terra: Terra is like Faire for custom goods | Y Combinator URL Source: Markdown Content: ### Terra is like Faire for custom goods Enter Terra, a new world of product building. Connect with suppliers, expedite sampling, and track production in real-time, removing the uncertainties of overseas manufacturing. Terra Founded:2024 Team Size:2 Location:New York ### Active Founders ### Lev Glushkovskii, Founder Previously co-founded SabersPro and bootstrapped the company from 0 to $10M+ in sales in 2.5 years, where he restructured the supply chain with factories increasing margins by 15% saving ~1M in profits. Dropped out from the Business, Technology, and Entrepreneurship program (BTE) at NYU Stern. ### Dylan McFarland, Founder In high school bootstrapped a furniture startup and in under a year went from selling on Craigslist to running a 40-person factory in Shenzhen. Took those insights and dropped out NYU Stern BTE to run Terra, a platform that streamlines supply chain creation for growing companies. ### Company Launches [### Terra 🌎: Instant supply chain creation]( **TL;DR -** ----------- Use [Terra]( as an all-in-one platform to seamlessly **develop, manufacture, and ship your products**. Discover pre-vetted suppliers and instantly create modular supply chains, all while having access to wholesale pricing regardless of order quantity. — **The Team** ------------ Hello all! We’re [Dylan]( and [Lev]( from Terra, and we’re _really_ passionate about making custom manufacturing accessible to everyone. **🚨 The Problem 🚨** --------------------- We’ve been working with suppliers globally since we were freshmen in high school. If you can think of a costly manufacturing or shipping mistake, we’ve likely already made it a dozen times over. But through our experiences building supply chains, we’ve routinely run into the same issues: ⚙️ Way too many moving parts ⚠️ The best factories have high minimum order requirements 👎 Current software is either too rudimentary or ridiculously complex **✅ The Solution ✅** -------------------- * Forget spending weeks juggling multiple supplier group chats, spreadsheets of shipping quotes, and financial docs. Terra automatically matches you with the right partners through a single dashboard, allowing you to launch products in minutes. * We aggregate demand based on similarity, ensuring you’ll get access to the best pricing and factories regardless of order size. * Our platform is completely modular. Build a supply chain entirely from our pre-vetted partners, onboard your existing suppliers, or use a combination of both. No constrictions, only flexibility. **🤝 Our Ask 🤝** ----------------- Are you or anyone in your network looking to launch/scale/streamline manufacturing processes? If so, hit us at [[email protected]](mailto:[email protected]), we’d love to get in touch. Currently looking specifically for anyone selling/manufacturing furniture or home goods! #### YC Sign Photo

Dodo

dodo

https://www.vetdodo.com/

San Francisco, CA, USA

Every third receptionist in specialty clinics leaves their job each year. At the same time, increased demand for specialized care has strained clinic resources. Facing understaffed and undertrained support, clients spend an average of 67% of their call time listening to hold music and navigating phone trees (“press 1 for x”). Frustrated by long wait times, many take their business elsewhere. Dodo is an AI receptionist for specialty clinics, such as veterinarians and dentists, fine-tuned on industry-specific knowledge and trained on protocols taken directly from receptionists’ manuals, which often span multiple binders. We handle front and back-office tasks, enabling clinics to provide 24/7 service. Dodo answers calls, sends medical records, and refills prescriptions - fully autonomously, with no human in the loop.

AI Receptionists for Specialty Clinics

Healthcare

Healthcare -> Healthcare IT

S24

Active

Early

https://www.ycombinator.com/companies/dodo

https://yc-oss.github.io/api/batches/s24/dodo.json

Title: Dodo: AI Workforce for Specialty Clinics | Y Combinator URL Source: Markdown Content: ### AI Workforce for Specialty Clinics Dodo builds an AI workforce for specialty clinics, such as veterinarians and dentists, fine-tuned on industry-specific knowledge and trained on protocols taken directly from employee manuals and playbooks, which often span multiple binders. Clinics hire Dodo's labor-as-a-software to handle front and back-office tasks, enabling clinics to provide 24/7 service. Dodo answers calls, sends medical records, and refills prescriptions—fully autonomously, with no human in the loop. Dodo Founded:2024 Team Size:3 Location:San Francisco ### Active Founders ### Dominik Moehrle, Founder Building AI agents for medical practices at Dodo. Previously studied Data Science @ Stanford, started beam.ai, and built & scaled 4 products in the healthcare space. When I'm not building, I'm teaching @ equala.org. ### Julius Hillebrand, Founder Building AI agents for medical clinics at Dodo. Stanford Engineer, Oxford PPE. Bootstrapped investment tool to $2m+ AUM, automated workflows for a large fintech, and led one of Oxford's biggest university organizations with 1000+ members. Former child actor. ### Tim Reinhart, Founder Building AI agents at Dodo. Studied Computer Science at Stanford and Robotics at ETH Zurich. Built autonomous drone systems at one of the worlds leading robotics labs and worked on mobile computer vision applications. ### Company Launches [### Dodo - AI Workforce for Specialized Healthcare Clinics 🦤]( Hey YC! We’re Dominik, Tim, and Julius from [Dodo]( Dominik and Julius grew up together, and we have known each other for 15 years. We were all in grad school together and just moved up from Stanford. **TL;DR: Dodo handles front- and back-office tasks for specialty clinic end-to-end** ------------------------------------------------------------------------------------ * Our AI agents answer calls, send medical records, and refill prescriptions—fully autonomously, with no human in the loop * We are starting with veterinary clinics, and are expanding to dental and physical therapy * Same-day cancellations trigger automatic outbound calls, filling availabilities on short notice * This means: No more hold music, phone trees (”press 1 for x”), or voicemail beeps (sorry if you like listening to jingles for minutes on end 🎶) ### **How you can help ❤️** * Know someone **who owns a specialty clinic, independently or through a larger corporate/PE?** We would be grateful for an introduction! * **Reach out **[**[email protected]**]( or book time on our **[**calendar**]( ### **Problem: Clinics can’t keep up with their admin work** * **🫥 Understaffed**: Clinics face labor shortages for administrative tasks, with staff feeling burned out due to long hours and inefficient workflows. * **🚆 Undertrained**: One in three medical receptionists leaves their job each year * **☎️ Phone-based**: Clinics handle up to 80% of their communication over the phone * **⏳ Result**: Bottlenecks in the front- and back-office, as clinics miss calls and clients. They also fail to fill in cancellations on short notice, decreasing efficiency. Callers spend 67% of their call time on hold ### **Solution: An AI Workforce for Specialty Clinics** * Handling all calls through verticalized AI agents trained on specialty knowledge and receptionists’ manuals (we’re talking literal binders 📚) * Interactions with internal software to update patient files, manage schedules, and execute necessary administrative tasks * Automatic outbound calls to fill missed appointments on short notice ### **👨‍👦‍👦 Meet the Team** If you’ve been to a specialty clinic recently, chances are you’ve seen us shadowing their front desk 🕵️‍♂️ **Julius** [[email protected]]( [ **Dominik** [[email protected]]( [ **Tim** [[email protected]]( [

Kastle

kastle

https://kastle.ai/

San Francisco, CA, USA

Kastle is an AI voice agent for mortgage servicing. We help servicers collect payments, verify new borrowers, and qualify new loan inquiries over the phone. Rishi and Nitish met at UIUC where they graduated with degrees in systems engineering and computer science. Rishi was the founding Product Manager for Redfin’s Mortgage Marketplace at 22 and grew the business from $3M to $6M in revenue in 9 months. Nitish patented Verkada’s motion search system and was the youngest tech lead at the company at 23. While building Redfin's Mortgage Marketplace Rishi realized traditional lenders didn't have the infrastructure to service online inquiries and were losing customers. Nitish and Rishi are building Kastle to help traditional financial services companies better interface with their customers.

AI voice agents for mortgage servicing

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/kastle

https://yc-oss.github.io/api/batches/s24/kastle.json

Title: Kastle: AI voice agents for mortgage servicing | Y Combinator URL Source: Markdown Content: ### AI voice agents for mortgage servicing Kastle is an AI voice agent for mortgage servicing. We help servicers collect payments, verify new borrowers, and qualify new loan inquiries over the phone. Rishi and Nitish met at UIUC where they graduated with degrees in systems engineering and computer science. Rishi was the founding Product Manager for Redfin’s Mortgage Marketplace at 22 and grew the business from $3M to $6M in revenue in 9 months. Nitish patented Verkada’s motion search system and was the youngest tech lead at the company at 23. While building Redfin's Mortgage Marketplace Rishi realized traditional lenders didn't have the infrastructure to service online inquiries and were losing customers. Nitish and Rishi are building Kastle to help traditional financial services companies better interface with their customers. Kastle Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Rishi Choudhary, Founder Rishi is the co-founder and CEO of Kastle ( Before Kastle, Rishi was the founding Product Manager for Redfin’s Mortgage Marketplace at age 22 and grew the business from $3M to $6M in revenue in 9 months. He also led Redfin's Search Experience team responsible for 80% of the company's traffic. Rishi has a BS in Systems Engineering from the University of Illinois at Urbana Champaign. At UIUC, Rishi ran the largest US student consulting club with 250 students and $500K ARR ### Nitish Poddar, Founder Nitish is the co-founder and CTO of Kastle (kastle.ai). Before Kastle, Nitish was a tech lead at Verkada, where he led development for a patented motion search system and built out the entire backend inference infrastructure. He also led the firmware delivery team on a new fleet management system and led engineering for Incident Management, Verkada's newest product line. Nitish graduated from UIUC with a degree in Computer Science. ### Company Launches [### Kastle - AI servicing for mortgage lending]( **✨** **TL;DR** [Kastle]( builds AI voice and SMS agents for mortgage lending. We help lenders collect payments, answer loan questions, and qualify new loan inquiries using an AI team member. **❌  The Problem** Mortgage servicers don’t have enough support agents to collect payments and answer customer questions. This leads to * Borrowers waiting in long queues just to make their payment or get a question answered. * Existing agents working long hours and multi-tasking to get through the queue. This leads to poor service and loss in repeat business for lenders. Since the call volume varies throughout the week, mortgage servicers are unable to scale their support agent team to meet the additional demand, leading to high average response times and unhappy loan investors. **🏰** **Enter Kastle** [Kastle]( automates calls to collect payments and answer escrow questions, using AI voice agents. This helps lenders provide 24/7 service to customers, drive upsells, reduce call center costs by 90%, and ensure no borrower ever waits in line to make a payment or answer a question about their loan. [Checkout Kastle's AI voice agent]( We seamlessly integrate with a lender’s mortgage servicing system and telephony system to perform actions, leave loan comments, and route calls back to live agents if needed. Lenders can use pre-built action blocks to configure the AI agent. Kastle also automatically scores each AI call against a lender’s quality framework to ensure regulatory compliance. **🤝 Team** We’re Rishi and Nitish and we met at UIUC while studying engineering and computer science. * [Rishi]( (CEO) led Redfin's Mortgage Marketplace team as a Product Manager and realized traditional mortgage lenders didn't have the infrastructure to service online inquiries and were losing customers. He also led Redfin's Search Experience team, which is responsible for 80% of the company's traffic, and is passionate about serving homeowners. * [Nitish]( (CTO) was a tech lead at Verkada, where he led development for a patented motion search system and built out the entire backend inference infrastructure. He also led the firmware delivery team on a new fleet management system and led engineering for Incident Management, Verkada's newest product line. **🙏 Ask** * **Please introduce us to any mortgage servicing or default management executives.** * \*\*If you’re a lender or sub-servicer looking to reduce costs and better retain customers, please [**book a demo here**]( or email us at \*\*[**[email protected]**](mailto:[email protected])

Propaya

propaya

https://www.propaya.com

San Francisco, CA, USA

Propaya is an automation application for rental property managers. By uploading a lease PDF, Propaya will handle everything from rent collection, utility transfer, to recommending local contractors and handymen.

AI-powered software to automate property management.

Real Estate and Construction

Real Estate and Construction -> Housing and Real Estate

S24

Active

Early

https://www.ycombinator.com/companies/propaya

https://yc-oss.github.io/api/batches/s24/propaya.json

Title: Propaya: AI-powered software to automate property management. | Y Combinator URL Source: Markdown Content: ### AI-powered software to automate property management. Propaya is an automation application for rental property managers. By uploading a lease PDF, Propaya will handle everything from rent collection, utility transfer, to recommending local contractors and handymen. Propaya Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Reader Wang, Founder Reader, CEO of Propaya, a former SpaceX engineer in the Starship Program with an MS from Stanford and a BS from the University of Michigan. Still actively managing over 70 rental properties for his family, Reader understands the firsthand pain points of property management. ### Jake Golas, Founder Jake, CTO of Propaya, a former full-stack software developer at Epic Systems, completed his BS in Computer Science in just three 1/2 years. ### Company Launches [### 🏠 Propaya - AI-powered Property Management Software]( **tl;dr:** Propaya simplifies the complexities of property management. Our platform automates routine tasks, provides real-time insights, and reduces the administrative load, empowering property managers to focus on what truly matters—growing their business. Available now at [Propaya.com]( **Meet the founding team behind Propaya - Jake & Reader!** **From left to right:** 🧠 [**Jake**]( is our analytical brain: CTO of Propaya, CS degree in 3½ years, former full stack software engineer at Epic Systems. Chat with him about #AI #Azure #dotnet **❤️** [**Reader**]( keeps an eye on the bigger picture: CEO of Propaya, 7+ years in real estate management and tech innovation, former SpaceX Engineer, and an MS from Stanford. Chat with him about #propertymanagement #proptech #realestate #propertymanagement * * * **Why is smart property management essential?** 🧐 Property management is evolving rapidly, and staying ahead requires adopting tools that simplify operations and drive growth. With increased demand for remote work and digital solutions, property managers are under pressure to do more with less—Propaya helps you meet that challenge. * * * We created Propaya because we understand the personal day-to-day challenges property managers face, from handling tenant requests and maintenance scheduling to managing finances and compliance. Our platform was built to automate these repetitive tasks, giving you back precious time and providing actionable insights to make informed decisions. * * * **The Propaya Advantage** 🎉 * **Automated workflows:** Say goodbye to manual data entry and hello to efficiency. Propaya handles everything from lease renewals to maintenance requests with ease. Just upload your lease PDF, and we will take it from there. * **Finance management:** Upload a receipt, and we'll keep your expenses organized and securely stored, forming detailed expense data sheets to make tax season easy. * **Automated calendar and alerts:** Stay on top of every task with our integrated calendar and alert system. Propaya automatically schedules important dates and sends timely reminders, ensuring you never miss a deadline or appointment. * **Remote management:** Manage your properties from anywhere, anytime, with our fully mobile-responsive platform. **Our Ask** 🙏 * **Try Propaya today!** Visit [ to explore our platform and see how it can transform your property management experience. * **Share this post!** Spread the word about Propaya with your network—whether they’re property managers, owners, or contractors, they’ll benefit from what we offer. * **Connect us with potential partners**: We’re looking to collaborate with property management consultants, real estate investors, realtors, and tech-forward property management individuals and firms. Reach out to us at [**[email protected]**](mailto:[email protected]) Quick blurb to share: _A Stanford-founded startup, Propaya, has launched a platform that automates property management, reduces administrative burdens, and provides real-time insights. Learn more at _[_ or contact _[[email protected]_](mailto:[email protected])_ for details._

Cerulion

cerulion

https://www.cerulion.com

Cerulion is middleware for robots - it abstracts away the process of interfacing with the motors and sensors. Visualization, data logging, and control can be easily achieved in a no-code GUI.

Cross-platform robot middleware with no-code GUI

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/cerulion

https://yc-oss.github.io/api/batches/s24/cerulion.json

Title: Cerulion: Open source operating system for robots | Y Combinator URL Source: Markdown Content: ### Open source operating system for robots Cerulion is an open source operating system for robots built by 2 MIT robotics PhDs. Companies like Amazon Robotics and Boston Dynamics AI Institute are using us to develop faster and ship more reliable robots than with ROS. With up to 1000x improvements in communications performance over state of the art, Cerulion is enabling the next generation of embodied AI. Cerulion Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Lakshay Sharma, Founder/CEO Lakshay is the cofounder/CEO of robotics middleware startup Cerulion. Prior to this, he studied computer engineering, astrophysics and robotics at the University of Pennsylvania, and researched robot motion planners at MIT. ### Se Hwan Jeon, Founder I'm the cofounder of Cerulion along with Lakshay Sharma. Cerulion is a software interface and abstraction layer to make controlling, debugging, and visualizing commercial robots accessible to everyone. Prior to YC, I studied mechanical engineering at the University of Notre Dame and MIT, specializing in robotics, optimal control, and reinforcement learning for legged platforms. ### Company Launches [### Cerulion 🤖 Fast, reliable, cross-platform robot middleware]( Robots are expensive and hard to develop/debug. [Cerulion]( solves the robot tooling problem to cut down development times by up to 10x. The Problem ----------- The incredibly high capital costs required for robotics, combined with the inherently non-deterministic nature of embodied intelligence, result in an unpredictable system where even attempts, let alone failures, are cost-prohibitive. **Robots are expensive and hard.** And yet, despite time on hardware being such a precious resource, roboticists today are absolutely wasting 90% of it fighting against unoptimized tooling. Some common complaints are: * Needlessly complex to install, maintain, and document * Incredibly steep learning curve for non-transferable skills * Poor observability into a robot's belief * Artificial non-determinism induced by slow or unreliable message-passing * Unreliable network protocols and IPCs * System breakdown due to poor inter-robot communication * Unreliable logging and cumbersome/nondeterministic log replay * Wasting precious field time parsing 1000s of lines of runtime errors whooshing past the screen And the list goes on and on. **Robot tooling today is broken. Unacceptably so.** Our Solution: Performance never looked this good! ------------------------------------------------- * **Packaged as an app, rather than a CLI tool:** Instead of shipping as a bunch of CLI tools to be installed, patched, and reinstalled, Cerulion is an app you can download on Windows, Mac, or Linux. It will smartly detect and handle any dependencies as needed. * **Intuitive, cross-platform visualizer:** The sleek, easy-to-use GUI is the main way of communicating with your robot in the field, which means you can run it with minimal resources on your favorite long-lasting laptop with any OS, instead of **_that_** RGB laptop with a gaming GPU running Linux. * **Minimal boilerplate and setup:** The way you program robots in Cerulion is designed to get out of your way, so you can focus on the planner and perception algorithms instead of worrying about the inter-process comms, IP addresses, or network routes. * **Fast and reliable communication:** We leverage cutting-edge Rust-based networking and IPC stacks like [Zenoh]( and [IceOryx2]( along with tried and tested C++ alternatives like [LCM]( and [IceOryx]( By ensuring communications don't fail silently and messages are delivered when expected, Cerulion eliminates nasty surprises in safety-critical scenarios. Meet The Team ------------- As robotics PhD students at MIT, [Lakshay]( and [Se Hwan]( have gathered years of experience (and frustration) working on robots, and want to see a world where you can take robot reliability for granted. Our Ask ------- Have experience with ROS or any other robot middleware? Did we miss something above that you really need? Send us a message at [[email protected]](mailto:[email protected])! We always love (and are happy to provide) warm intros to people working on robots and automation. And of course, if you need help with your robot system, or are curious about any of this, we're happy to chat! P.S. — Please star our [GitHub repo]( #### Company Photo

Dataleap

dataleap

https://dataleap.ai/

San Francisco, CA, USA

Dataleap is a GenAI-powered market research platform for management consulting firms and corporate strategy departments. Think Perplexity for Consultants. We combine the natural language interface from chatGPT with trusted, verifiable market data from best-in-class data providers. We do the heavy lifting of aggregating all data sources into a single interface. We optimize the whole market research workflow from finding the right information, managing the internal knowledge base as well as sharing information with team members. We already work with leading consultancies like BCG and industry giants like Siemens Energy and Uniper.

Perplexity for Consultants

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/dataleap

https://yc-oss.github.io/api/batches/s24/dataleap.json

Title: Dataleap: Perplexity for Consultants | Y Combinator URL Source: Markdown Content: ### Perplexity for Consultants Dataleap is a GenAI-powered market research platform for management consulting firms and corporate strategy departments. Think Perplexity for Consultants. We combine the natural language interface from chatGPT with trusted, verifiable market data from best-in-class data providers. We do the heavy lifting of aggregating all data sources into a single interface. We optimize the whole market research workflow from finding the right information, managing the internal knowledge base as well as sharing information with team members. We already work with leading consultancies like BCG and industry giants like Siemens Energy and Uniper. Dataleap Founded:2023 Team Size:2 Location:San Francisco ### Active Founders ### Jan Damm, Founder Co-founder and CEO at Dataleap. Before Dataleap, I worked as an APM in the Google search team as well as a consultant at McKinsey and the Siemens AI Lab. ### Jan-Hendrik Ruettinger, Founder Cofounder and CTO at Dataleap. Before Dataleap, I worked as an AI researcher at Oxford University and Volkswagen. I also had a brief stint in consulting at BCG. ### Company Launches [### Dataleap - Perplexity for Consultants]( **_TL;DR -_** [_Dataleap_]( is a GenAI-powered market research platform for management consulting firms and corporate strategy departments. **Think Perplexity for Consultants.** _Our secret sauce is combining the natural language interface from Perplexity with trusted, verifiable market data from best-in-class market research firms and the company’s internal resources. We already work with two of the three tier 1 consultancies (MBB) and industry giants like Uniper and Siemens Energy._ [_Try it out here!_]( **Problem:** ------------ * AI search tools like **Perplexity or ChatGPT are not cutting it for professionals analyzing markets**. The answers are either hallucinated or using low-quality sources. * Consultants & corporate strategy departments need to work with **paywalled data** curated from expert market research firms like WoodMackenzie, Rystad, or S&P Global. * They do not want to interact with 15 different platforms or badly implemented AI chats but with one platform that is close to them and aggregates everything. **Solution:** ------------- 1. Dataleap provides a **powerful AI Copilot** that aggregates and searches through all the relevant market data that strategy teams need to look at. 2. We always **highlight the exact sentence** in the original data where the information is coming from so that you can trust it and make real business decisions based on it. 3. You can also **upload your proprietary information** to have the full picture of all your available information within Dataleap. 4. You can not just search through the documents but also conduct meta-studies of reports or extract your own datasets from documents with our **document matrix** **About us:** ------------- [Jan-Hendrik Rüttinger]( and I ([Jan-Hendrik Damm]( met during our studies at CDTM in Munich 5 years ago and shared an apartment ever since. We instantly knew that we would start a company together one day, not just because we had the exact same uncommon name but because we just clicked. I went on to work in consulting at McKinsey and as an APM at Google in the search team. He decided to pursue machine learning research at Oxford. Last year, we decided it was time to make the leap, moved to the US, and started Dataleap. **Ask:** -------- **Please introduce us to strategy and market intel teams at big energy companies** (energy is the first industry we started with), e.g., Dominion Energy, RWE, BP, Shell, GE Vernova. → Email address: [**[email protected]**](mailto:[email protected]) #### YC Sign Photo

Freestyle

freestyle

https://www.freestyle.sh

San Francisco, CA, USA

We are building an abstract cloud for JavaScript engineers. We replace databases, message queues, streaming infrastructure, authorization, and more with open-source JavaScript packages that developers can compose into world-class software. To modularize this capability in JavaScript, we developed a new distributed runtime, hosted by our cloud.

Build exceptional infrastructure with just JavaScript.

B2B

B2B -> Infrastructure

S24

Active

Early

https://www.ycombinator.com/companies/freestyle

https://yc-oss.github.io/api/batches/s24/freestyle.json

Title: Freestyle: Build exceptional infrastructure with just JavaScript. | Y Combinator URL Source: Markdown Content: ### Build exceptional infrastructure with just JavaScript. Ben and Jacob met in Calculus class in 10th grade, and tried founding a startup in high school — it didn’t go anywhere, but it taught them a ton. During high school, Ben lead the frontend team for the World Health Organization’s COVID-19 App Team. After high school, Jacob went to Apple and Ben went to college. When there was an opening on his team, Jacob had Ben apply, and Ben left UChicago to go work at Apple. They spent about 2 years at Apple as Senior Engineers in the Conversational Engineering Department, before leaving to found Freestyle. Freestyle is directly inspired by what they saw as their failures of the tech they worked with at Apple. Freestyle Founded:2024 Team Size:5 Location:San Francisco ### Active Founders ### Benjamin Swerdlow, Founder I learned to code in elementary school. In high school I served as a Code Owner and technical lead for the World Health Organization COVID-19 App Collective Team. I did a year of college, then took a job as a consultant for Apple. In 3 months I became a senior. I left Apple to found Freestyle. ### Jacob Zwang, Founder/CTO I dropped out of college and became an Apple senior full-stack engineer in Conversational Engineering at 19. During the year and a half I worked there, I delivered multiple projects to production at scale. I love building products that inspire and delight customers, but I'm frequently frustrated by poor back-end developer tools. My co-founder and I are building a new cloud and backend workflow so that we and others can build exceptional software in a fraction of the time. ### Company Launches [### Freestyle: A new way to build fullstack TypeScript apps]( Today, most new apps are written with Fullstack TypeScript, but the tooling around them is still built as though the frontend and backend are separate codebases. We built an open source framework, runtime, and cloud that acknowledges this and removes these unnecessary layers from the development process. **What?** Any JavaScript class can be marked as [@cloudstate]( and: 1. The data in it is now persisted. You query the data with regular JavaScript. 2. The methods on it can now be called from the front end. No Express/API Routes/tRPC needed! **Coolest Side Effect:** By building our tooling around Fullstack TypeScript, we enable **full-stack feature packages,** a way for JavaScript developers to package any feature they build and share it in a way that it can be embedded in other projects. Think **npm install mui**, except it’s **npm install chat**, or **npm install auth**. Imagine a world where every third-party SaaS provider today can be an open source package that runs in your application. **Our Ask:** **Build with us!** Contribute to our [Open Source Repository,]( or try [deploying an app with us]( ### Other Company Launches ### Freestyle: TypeScript for EVERYTHING A fundamental shift in how we approach web development [Read Launch ›](

Flyflow

flyflow

https://flyflow.ai

San Francisco, CA, USA

"I don’t know how I would have found these leads myself." - YC Founder Flyflow enables companies to find prospective customers using AI workflows instead of drowning in tables. Try out our product at app.flyflow.ai. How are we different? (1) Higher quality leads: get to a list that’s so relevant, you’ll want to spend time getting introduced to people there, instead of sending another 1,000 cold emails (2) Incredibly easy to set up: get up and running with a list of prospects in hand in under 10 minutes (3) Integrate the best of different sales tools all in one place: no need to switch between different tabs for web research, filtering by company size, and finding prospect LinkedIns

The easiest way to find high-quality sales leads

B2B

B2B -> Sales

S24

Active

Early

https://www.ycombinator.com/companies/flyflow

https://yc-oss.github.io/api/batches/s24/flyflow.json

Title: Flyflow: The easiest way to find high-quality sales leads | Y Combinator URL Source: Markdown Content: ### The easiest way to find high-quality sales leads "I don’t know how I would have found these leads myself." - YC Founder Flyflow enables companies to find prospective customers using AI workflows instead of drowning in tables. Try out our product at app.flyflow.ai. How are we different? (1) Higher quality leads: get to a list that’s so relevant, you’ll want to spend time getting introduced to people there, instead of sending another 1,000 cold emails (2) Incredibly easy to set up: get up and running with a list of prospects in hand in under 10 minutes (3) Integrate the best of different sales tools all in one place: no need to switch between different tabs for web research, filtering by company size, and finding prospect LinkedIns Flyflow Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Carl Cortright, Founder Hi there! My name is Carl. I'm an engineer and CTO of flyflow, an AI lead generation platform. In my free time I enjoy going to hackathons, hiking, skiing and generally being outside. ### Pritee Tembhekar, Founder Pritee is co-founder and CEO of Flyflow. Flyflow is the backend to power conversational voice agents, from customer support, to lead qualification, to candidate screening, and SMB scheduling. Pritee was previously Head of Product at Lithic (card-issuing APIs and Privacy.com consumer brand). Before that, she held product roles at Plaid and Coinbase. She started her career at BCG and Google and holds degrees from MIT and HBS. ### Company Launches [### Flyflow: The easiest way to find high-quality leads]( **TLDR;**  [Flyflow]( enables companies to find prospective customers using AI workflows instead of drowning in tables. Check it out (and find companies in <5 minutes). [Sign up and try our product for free]( yes free! **Impact** > I don't know how I would have found these leads myself. > > \-Varsha Ramesh Walsh, CEO, Offstream > Flyflow did the best job possible finding companies in what is probably a non-existent niche. > > \-Peter Marler, CEO, Unit IQ > Flyflow has helped us surface the signals that matter most in customer qualification. Our lead list is already better than what we've been working off before. > > \-Ankit Singhal, CEO, Arctic Capture Over the last three weeks, we’ve worked with ~30 companies in our batch and dozens more at larger companies. Our customers have significantly improved their lead sourcing process, now focus on higher-quality leads, and will soon have increased sales. Join the fun! **Demo** Imagine if you could find your ideal prospects quickly, automating the manual web research and filtering that you’re doing now. You would never have to learn how to use Apollo, Clay, Linkedin Sales Navigator, Instantly, or other sales tools. Our system includes lookalike search, meta-data filtering, and concise lead notes. **Your team could be finding high-quality leads, today** We'd love for you to [try our platform]( see you improve your sales pipeline, and hear any feedback you have. Mention us: If you like what you see, mention us in Slack groups/social where sales leaders, SDRs, or startup founders hang out. **Wait, but why** Our journey began with a tale as old as time: we were struggling to sell enterprise contracts for our AI voice product. Even finding companies who perfectly fit our Ideal Customer Profile (ICP) required hours of manual work to research each company to see if the size, product, and team were a good fit for us. THEN we had to get a meeting. We even stopped building so that we could “focus on GTM” ☠️. None of the sales tools we tried solved this for us. This experience led us to develop Flyflow. Our vision is to make high-quality lead sourcing accessible to everyone, without feeling like you need to learn a new set of tools. **The Team** [Carl]( and [Pritee]( love building products that make really annoying tasks go away. Most recently, Pritee was Head of Product at Lithic (launch credit/debit cards insanely fast), and before that was on the product team at Plaid (link your bank account to any app in seconds, not days). Carl was building Coherent, making Web3 infrastructure so that builders could integrate web3 data (in minutes instead of months). They met at Coinbase in 2018; those were the days. And so are these. ✨

Parahelp

parahelp

https://parahelp.com/

San Francisco, CA, USA

Parahelp is an AI support agent for software companies that resolves tickets end-to-end by using the same tools as human agents. For example, when a customer requests a refund, Parahelp can automatically look up subscription details in Retool, process the refund in Stripe if eligible, and respond to the customer in Intercom.

The AI support agent used by leading AI companies

B2B

B2B -> Operations

S24

Active

Early

https://www.ycombinator.com/companies/parahelp

https://yc-oss.github.io/api/batches/s24/parahelp.json

Title: Parahelp: The AI support agent used by Perplexity and Captions.ai | Y Combinator URL Source: Markdown Content: ### The AI support agent used by Perplexity and Captions.ai Parahelp is an AI support agent for software companies that resolves tickets end-to-end by using the same tools as human agents. For example, when a customer requests a refund, Parahelp can automatically look up subscription details in Retool, process the refund in Stripe if eligible, and respond to the customer in Intercom. Parahelp Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Anker Ryhl, Founder Co-founder of Parahelp. I began programming at 13, where I also founded my first company. Throughout high school, I built consumer apps with my also current co-founder, Mads. Now, we’re building Parahelp - The AI support agent for software companies. ### Mads Liechti, Founder Co-founder of Parahelp. Throughout high school, I built 4 consumer apps with my also current co-founder, Anker. Now, we’re building Parahelp - The AI support agent for software companies. ### Company Launches [### Parahelp - The AI support agent for software companies]( **TL;DR** - [Parahelp]( is an AI support agent for software companies that resolves support tickets end-to-end. **For example:** When a customer requests a refund, Parahelp can automatically look up subscription details, process the refund in the payment system if eligible, and respond to the customer. **Problem** Most CX teams have tried a chatbot solution that didn't move the needle in production. This is because resolving tickets end-to-end, such as billing inquiries and technical issues, requires using multiple tools to look up customer data (subscription status in RevenueCat, platform details in Retool, etc.) and execute relevant actions (process refunds in Stripe, create bug reports in Linear, etc.). **Solution** We've built Parahelp as an AI agent that works with CX teams' existing tools to reliably resolve support tickets end-to-end in production. Parahelp can be highly customized to the team’s knowledge and processes and implemented in days, not months. Our state-of-the-art agent infrastructure enables our AI support agent to: * Be human-level empathetic and engaging * Quickly adapt to product and company changes * Take actions in tools like Stripe, Retool, and Linear * Work directly in Intercom and Zendesk **Results** We're already helping leading CX teams at Captions and Perplexity. Together with Captions, we’ve created a [case study]( about how Parahelp was resolving 46% of all their support tickets within 7 days of implementation. **Ask** We'd love to chat if you're a software company looking to implement an AI solution for your support operations! Book a demo with us here: [

Opslane

opslane

https://opslane.com

San Francisco, CA, USA

Opslane speeds up incident resolution by automating work that is performed by short staffed SRE teams. We do this by stitching together relevant context (logs, metrics, traces, code changes and runbooks) to help on-call engineers figure out what's going on.

AI SRE to speed up incident response

B2B

B2B -> Infrastructure

S24

Active

Early

https://www.ycombinator.com/companies/opslane

https://yc-oss.github.io/api/batches/s24/opslane.json

Title: Opslane: AI SRE to speed up incident response | Y Combinator URL Source: Markdown Content: ### AI SRE to speed up incident response Opslane speeds up incident resolution by automating work that is performed by short staffed SRE teams. We do this by stitching together relevant context (logs, metrics, traces, code changes and runbooks) to help on-call engineers figure out what's going on. Opslane Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Amaresh Ray, Co-founder Co-founder at Opslane. I was a PM on Atlassian’s Jira Service Management, which was the fastest growing product in Atlassian’s history, growing from 20,000 to over 50,000 customers in 3 years. ### Abhishek Ray, Founder Co-founder @ Opslane. Abhishek was an early engineer at Robinhood and Quora. He was at Robinhood for 5 years where he helped scale infrastructure from ~1M to 20M+ customers. He saw firsthand how challenging incident response can be when critical information is spread across 6 different observability tools. ### Company Launches [### Opslane – Making on-call suck less]( **TL;DR** Our [Slack bot]( classifies your Datadog alerts and flags noisy alerts, reducing alert fatigue. Additionally, we generate insights based on all your alerts to give you a picture of your overall alert hygiene. ### The Problem **Picture this:** It's 3AM, and you're woken up by a page. It's an error you've seen before, and you know it wasn't worth waking up for. Anyone who's been on-call knows the infamous middle-of-the-night unactionable alert all too well. Even with modern tools, it's still tough to determine when something has actually gone wrong, the user impact, and how to fix it quickly. **Why are existing solutions not cutting it?** * Lack of context - it's really hard to understand the customer and business impact from looking at an alert. * Alerts are noisy and not actionable. * Debugging requires context switching between multiple siloed tools. ### Our solution We're starting with our open source Slack bot that reduces alert noise. Once you add our bot to your Slack channel, we pull your alert history from Datadog. We analyze your alert history across various parameters: * Alert frequency * How quickly have the alerts resolved in the past * Alert priority * Alert response history Based on these, we classify your alert as actionable or noisy. In addition, we generate a report on your team's overall alert health, pinpointing which noisy alerts are prime candidates for tuning or disabling. Our vision is to stitch together all the relevant context engineers need to debug incidents in one place and automate root cause analysis. ### Team We ([Amaresh]( and [Abhishek]( have firsthand experience dealing with the challenges of on-call and alert fatigue at companies like Robinhood and Atlassian. With Opslane, we’re reducing the alert fatigue that overwhelms on-call engineers and leads to millions in lost productivity and downtime. ### Ask 🔥 Are noisy alerts a hair-on-fire problem for your company or someone you know? We'd love to hear from you - email [[email protected]](mailto:[email protected]) or [schedule a call]( with us. ⭐ Please star us on Github: [

Synnax

synnax

https://synnaxlabs.com

San Francisco, CA, USA

We build software that helps hardware teams quickly and clearly evaluate how a component (engine, battery, valve, etc.) or process (assembly line, manufacturing plant) performs. We prevent mistakes, reduce costs, and accelerate time-to-market. Our platform unifies unifies industrial control with sensor data storage and automated post-processing, delivering an end-to-end operations and observability stack. To put it comparatively, take LabVIEW, InfluxDB, and Grafana, wrap them into a single package, add a bit of magic, and you get Synnax.

Control & Sensor Data Processing for High-Cadence Hardware Teams

B2B

B2B -> Infrastructure

S24

Active

Early

https://www.ycombinator.com/companies/synnax

https://yc-oss.github.io/api/batches/s24/synnax.json

Title: Synnax: Unified Hardware Control and Data Processing Software | Y Combinator URL Source: Markdown Content: ### Unified Hardware Control and Data Processing Software We build software that helps hardware teams quickly and clearly evaluate how a component (engine, battery, valve, etc.) or process (assembly line, manufacturing plant) performs. We prevent mistakes, reduce costs, and accelerate time-to-market. Our platform unifies unifies industrial control with sensor data storage and automated post-processing, delivering an end-to-end operations and observability stack. To put it comparatively, take LabVIEW, InfluxDB, and Grafana, wrap them into a single package, add a bit of magic, and you get Synnax. Synnax Founded:2024 Team Size:3 Location:San Francisco ### Active Founders ### Emiliano Bonilla, Founder Co-Founder & CEO at Synnax. I'm classically trained in Aerospace Engineering and have a passion for building and selling great software to improve the way teams develop hardware. I led a team in the development of data processing tools for the first launches of Firefly Aerospace's Alpha rocket, one of the few new-space companies to repeatedly reach orbit. I've also worked on automous electric vehicle prototypes and developed banking software for European startups. ### Patrick Dotson, Founder Co-Founder & COO at Synnax. I worked on propulsion systems for the Starship program at SpaceX and have experience in both hardware and software engineering. I graduated from the University of Michigan with a degree in aerospace engineering. ### Elham Islam, Founder Co-Founder & CTO at Synnax. I wrote flight software for the Mars Sample Return Lander at NASA JPL and developed embedded control systems for laser communication at MIT Lincoln Laboratory. I studied computer engineering at the University of Michigan. ### Company Launches [### Synnax - The software infrastructure for operating & observing complex hardware]( ### **Summary** Teams operating hardware from rockets to manufacturing lines use outdated, unintuitive control software that disregards the importance of rapid, quality data processing. This leads to unexpected downtime, disastrous anomalies, and delayed development. [Synnax]( delivers a modern mission control stack with integrated data storage & processing. Wrap [LabVIEW]( [Snowflake]( and [Grafana]( into a single package, tune it for hardware, and you get Synnax. ### **The Problem** We’re a group of Aerospace Engineers hailing from NASA JPL, SpaceX, and Firefly Aerospace. (From left to right: [Patrick Dotson,]( [Elham Islam]( and [Emiliano Bonilla]( We’ve seen the best engineers blow up multi-million dollar systems because their mission control software didn’t give them a comprehensive picture of how their hardware was behaving. A director of test & launch operations once told our team: > It’s amazing what people guess at just because loading data is so hard To put their quote in context, here’s what the leading aerospace control software, LabVIEW, shows as a UI screenshot on their landing page: Remember [Scratch]( or [Lego Mindstorms]( from elementary school? LabVIEW is kind of like that, except people use it to fly humans to space. LabVIEW and its cross industry counterparts (such as [AVEVA Scada]( and [Rockwell FactoryTalk]( were first developed in the 20th century and have evolved little since. These tools come with: * Outdated, unintuitive user interfaces * High price tags * Closed ecosystems with no integrations * Poor documentation * Expensive certifications **Most importantly, they disregard the importance of tightly integrated and automated data processing that drives the innovation of today’s software giants.** Most solutions spit out massive CSV files or mindlessly write data to an SQL table. They’re disorganized, difficult to locate, inaccessible to most engineers, and are challenging to integrate with analysis tools. ### **Our Solution - The Product** At Synnax, we’re building a modern alternative to tools like LabVIEW by developing tightly integrated hardware control, sensor data storage, and automated post-processing software. Here’s our UI: We deliver: * Intuitive interfaces for humans to manually operate everything from automotive test stands to manufacturing lines. * Tools for categorizing and reviewing historical data. * Automated post-processing and report generation pipelines. * Direct integration with existing sensor systems so teams don’t need to buy new hardware. * A simple API for writing automated control sequences in Python, C++, and TypeScript. It’s this easy to open a valve based on temperature in Synnax: Here’s a comparable sequence in LabVIEW: ### **Our Solution - The Tech** We’ve built one of the fastest time-series database engines available. Synnax can stream millions of samples per second in real time, making it possible to control systems with extremely tight timing requirements, such as starting a rocket engine. ### **Our Solution - The Impact** By bringing together hardware control and sensor data processing, organizations operating large, complex systems tighten the feedback loop between recording sensor data and using it to make actionable decisions. With Synnax, teams: * Reduce risk by detecting and reacting to anomalies in real time. * Spend less time developing in-house data processing solutions. * Reduce the manual labor involved in extracting insights from sensor data. ### **Our Asks** Do you work on hardware and need updated control software? Do you know someone who does? Let us know at [[email protected]](mailto:[email protected]). ### Hear from the founders #### How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?) During his time in industry, Emiliano witnessed outdated control software and bad data review processes lead to multi-million dollar failures of critical path hardware. Frustrated with proprietary file formats, slow data pipelines, and unintuitive, complex tools like LabVIEW, he set out to built a modern control and data processing stack for hardware.Patrick and Elham were Emiliano’s first beta testers on the University of Michigan’s student rocketry team. They played an essential role in early iterations of the product, giving feedback through experience gained during their internships at SpaceX, NASA JPL, and MIT’s Lincoln Labs. Over time, their conviction in the problem Synnax is solving and the quality of the product grew. In early 2024, they joined forces as co-founders.

Arva AI

arva-ai

https://www.arva-ai.com

San Francisco, CA, USA

Every time a business onboards onto a bank or fintech a human analyst has to conduct manual document verification. We replace those human KYB (know your business) analysts with AI agents that handle cases in seconds, not days. Currently, 30% of know your business (KYB) cases result in manual review work. With Arva's AI agents the whole end-to-end onboarding experience is automated. Instant onboarding whilst reducing operational costs by 80%. Our end goal is to have a whole suite of AI workers that can handle manual compliance work for banks and fintechs. A $24B opportunity in the US alone.

AI agents for instant business verification

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/arva-ai

https://yc-oss.github.io/api/batches/s24/arva-ai.json

Title: Arva AI: AI agents for instant business verification | Y Combinator URL Source: Markdown Content: ### AI agents for instant business verification Every time a business onboards onto a bank or fintech a human analyst has to conduct manual document verification. We replace those human KYB (know your business) analysts with AI agents that handle cases in seconds, not days. Currently, 30% of know your business (KYB) cases result in manual review work. With Arva's AI agents the whole end-to-end onboarding experience is automated. Instant onboarding whilst reducing operational costs by 80%. Our end goal is to have a whole suite of AI workers that can handle manual compliance work for banks and fintechs. A $24B opportunity in the US alone. Arva AI Founded:2024 Team Size:2 Location:San Francisco ### Active Founders ### Rhim Shah, Founder/CEO Formerly led the FinCrime product team at Revolut Business and multiple product teams at Jobandtalent. Prior to that an APM intern at Google. Studied Engineering at the University of Oxford. ### Oli Wales, Founder/CTO Formerly lead product engineer at Opvia (YC S20), before that a full-stack engineer at Iventis, Chimnie and The Trade Desk. Studied Computer Science at the University of Cambridge. ### Company Launches [### Arva AI — AI agents for instant global KYB onboarding]( **✨ TL;DR** ----------- [Arva AI]( ([LinkedIn]( makes onboarding onto a bank/fintech as a business 100x faster and significantly cheaper. We've built a highly compliant AI agent that instantly handles all low/medium risk manual KYB work in real-time as a customer is onboarding, cutting reviews to seconds and ops spend by 80%. **❌ The Problem** ----------------- When a business onboards with a bank or fintech, a human analyst has to conduct manual verification over 40% of the time. **Why is this human review stage a problem?** * Longer onboarding times cause drop-off from the onboarding funnel (in some cases, up to 50%!). This is a huge revenue loser. * Large operational overheads to run compliance teams— usually in the millions for large fintechs. * Humans make mistakes and deviate from procedure. **💡 Enter Arva AI** -------------------- Arva AI’s agents completely replace human analysts, resulting in instant onboarding for low/medium-risk cases. Arva is a modular solution that can combine with preexisting compliance stacks or completely replace them as an end-to-end platform. Arva is capable of conducting advanced web due diligence on a business entity, document verification, fraudster checking, handling the back and forth with customers, and much more. **See it in action** [ **🙏 Asks** ----------- If you’re a fintech who has to conduct KYB onboarding, we’d love to discuss how we can make your onboarding times 100x faster and save you a bunch of money in the process. Book a demo [here]( or drop us an email: [[email protected]](mailto:[email protected]). Or if you’re connected to the fintech space and there is someone you can introduce us to, please reach out! **🦸 The team** --------------- We’re Rhim (Oxford MEng) and Oli (Cambridge BSc), and we’ve been building products together for a while! * [Rhim]( (CEO) formerly led the FinCrime product team at Revolut Business, where he saw the extent of the manual workload during business onboarding. He also led multiple product teams at Jobandtalent. Prior to that, he was an APM intern at Google. * [Oli]( (CTO) was formerly the lead product engineer at Opvia (YC S20). Before that, he was a full-stack engineer at Iventis, Chimnie, and The Trade Desk. #### Company Photo

Kontigo

kontigo

https://kontigo.lat/

San Francisco, CA, USA

We’re building Kontigo, a USDC Smart Neobank for Latinos in the U.S. and Latin America. We launched our Peer-to-Peer onramp 20 days ago and we’ve already had $540k in deposits, 10k active users and a +300k waitlist. The founding team has built and scaled fintechs in Latam, and it’s ex-Venmo, Nubank, Rappi, MercadoLibre, Ualá, Platzi, and Yuno. We’re already backed by DST Global, Soma Capital, Pioneer Fund, Transpose + 10 YC alums. Sending money to Latin America is just as painful as holding it due to the region's fragmentation into 33 countries with 39 different currencies. On one hand, cross-border payments to Latam are absurdly expensive (up to 20% per transaction). Conversely, over the past decade, currency depreciation across the region has become unsustainable. (some countries surpassing the trillion percent ). Kontigo solves this with a USDC global wallet & a BTC savings account. - Like Venmo, but on-USDC. - Like Zelle, but for cross-border payments. No limits. - Like Nubank & Revolut, but without inflationary currencies. Bitcoin-backed. User wallets are connected to an AI-private banker on WhatsApp to execute international USDC transactions.

USDC-Smart Neobank for Latinos.

Fintech

Fintech -> Banking and Exchange

S24

Active

Early

https://www.ycombinator.com/companies/kontigo

https://yc-oss.github.io/api/batches/s24/kontigo.json

Title: Kontigo: USDC-Smart Neobank for Latinos. | Y Combinator URL Source: Markdown Content: ### USDC-Smart Neobank for Latinos. We’re building Kontigo, a USDC Smart Neobank for Latinos in the U.S. and Latin America. We launched our Peer-to-Peer onramp 20 days ago and we’ve already had $540k in deposits, 10k active users and a +300k waitlist. The founding team has built and scaled fintechs in Latam, and it’s ex-Venmo, Nubank, Rappi, MercadoLibre, Ualá, Platzi, and Yuno. We’re already backed by DST Global, Soma Capital, Pioneer Fund, Transpose + 10 YC alums. Sending money to Latin America is just as painful as holding it due to the region's fragmentation into 33 countries with 39 different currencies. On one hand, cross-border payments to Latam are absurdly expensive (up to 20% per transaction). Conversely, over the past decade, currency depreciation across the region has become unsustainable. (some countries surpassing the trillion percent ). Kontigo solves this with a USDC global wallet & a BTC savings account. - Like Venmo, but on-USDC. - Like Zelle, but for cross-border payments. No limits. - Like Nubank & Revolut, but without inflationary currencies. Bitcoin-backed. User wallets are connected to an AI-private banker on WhatsApp to execute international USDC transactions. Kontigo Founded:2023 Team Size:10 Location:San Francisco ### Active Founders ### jesus alberto Castillo ferrer, CEO & Co-founder I'm a Venezuelan founder obsessed with closing the financial inclusion gap for 1.7 billion humans. - Previously built the largest micro-lending platform in Venezuela, reaching over 500k SMBs in low-income communities with no financial history, implementing Muhammad Yunus methodology powered by ML. - Co-founded the first LLM marketplace in LatAm using micro-payments instead of subscriptions. - Founder at Bicuye, the most exotic Agave-Based Liquour with an Ultra-Premium membership club. jesus alberto Castillo ferrer [Kontigo]( ### Gino Guatavita, Founder Co-Founder & CTO at Kontigo. Tackling one of the biggest challenges in the region, access to credit and the financial system. Currently 500M+ people on Latam are under-banked, we are solving this with alternative data, AI and Web3. ### Company Launches [### Kontigo - A USDC-smart neobank for Latinos]( **TL;DR**; We're connecting all Latinos across the U.S. and Latin America through a single banking app that is global, instant, and stable. Hi everyone, we are Jesus & Gino, and we’re building [Kontigo]( —**A USDC-smart neobank for Latinos.** (With a Bitcoin piggy bank). **The real problem:** * The fundamental problem behind high fees, underbanked users, and failed currencies, both in Latam and the US, is precisely the same: the lack of a stable currency with global access and minimum government interference. **The consequence of the problem:** * The U.S. is home to 65 million Latinos, representing the fastest-growing population in the country; despite this, 35% of them are underbanked. * Each year, Latinos in the U.S. send over $170 billion to Latin America, often incurring fees as high as 20%. * Once the money reaches Latam, recipients may incur additional fees of up to 10% to convert and cash out. * In contrast, Latin America has a population of 700 million, with 20% of households relying on remittances for income and 75% of the population being underbanked. * Latin America's financial fragmentation worsens the problem, with 33 countries using 39 different currencies. Over time, all of these currencies have failed due to devaluation, inflation, and poor monetary policies. **Kontigo offers a USDC global wallet with a BTC savings account.** * Like Venmo, but on USDC. * Like Zelle, but for cross-border payments. No limits. * Like Nubank & Revolut, but without inflationary currencies. Bitcoin-backed. Made for real-life use cases—no crypto BS. User wallets are connected to an **AI-private banker on WhatsApp** for their daily needs. * Execute international USDC transactions. * Build-Credit. Build Wealth. * Plan your BTC retirement. **We have brought together a “founder mode” dream team!** * +10 years building in fintech. * CEO prev built the largest micro-lending app in Venezuela. * CTO prev built and scaled 2 fintech companies in Colombia, Mexico and Ecuador. * CPO prev built A WhatsApp co-pilot for business owners. Ex-Platzi and HolaCasa. (YC companies). * COO prev built an LLM marketplace for Latam using micro-payments instead of subscriptions. * Head of design prev was founding team at Rappi, Grability, Imaginamos and Yuno. * Head of branding prev built an ultra-premium hand-crafted agave spirit brand. * Head of Web3 prev built an on-chain grocery delivery app for Colombia. * Head of engineering prev was tech lead at Venmo app development. (+Built a B2B payment startup in Mex). * Head of Content prev developed Ualá and Mercado Libre marketing strategy. Kontigo is: * Already backed by DST Global, Soma Capital, Pioneer Fund, Transpose, HF0, Founders, Inc. Orange DAO, Optimism. +10 YC Alum. * Profitable. * +2 year runway. We are gonna make it because we have done it before. But this time, we're aiming for the moonshot. Help us unlock universal wealth :) Ask: * Have a Latino friend? Share the [app]( with friends who need it. * Do you know any business, startup, or company that makes cross-border payments or needs a product like ours? Connect us! Contact at [[email protected]](mailto:[email protected]) **Stable Currency, Global Access.** ### YC S24 Application Video

Paasa

paasa

https://paasa.com

Gurugram, HR, India

Paasa is the global investing platform for aspiring HNIs. We simplify access to global markets from emerging economies, while offering an AI advisor and robo-investing. ‘Pre-HNIs’ and HNIs in countries like India want to diversify their wealth internationally. Most customers hesitate due to lack of guidance on foreign investments, tax laws that change every year, and slow outflows from local banks. These barriers keep global investing out of reach, reserved for the top 0.001% who can afford wealth managers and family offices. We’re changing that. Our vision is to bring all the opportunities of the 0.001% without any hassle.

Give HNIs in emerging markets a global investment portfolio

Fintech

Fintech -> Asset Management

S24

Active

Early

https://www.ycombinator.com/companies/paasa

https://yc-oss.github.io/api/batches/s24/paasa.json

Title: Paasa: Give HNIs in emerging markets a global investment portfolio | Y Combinator URL Source: Markdown Content: ### Give HNIs in emerging markets a global investment portfolio Paasa is the global investing platform for aspiring HNIs. We simplify access to global markets from emerging economies, while offering an AI advisor and robo-investing. ‘Pre-HNIs’ and HNIs in countries like India want to diversify their wealth internationally. Most customers hesitate due to lack of guidance on foreign investments, tax laws that change every year, and slow outflows from local banks. These barriers keep global investing out of reach, reserved for the top 0.001% who can afford wealth managers and family offices. We’re changing that. Our vision is to bring all the opportunities of the 0.001% without any hassle. Paasa Founded:2024 Team Size:2 Location:Gurugram, India ### Active Founders ### Nitish Sahni, Founder Building the next-generation wealth manager for all emerging economies, starting with India. ### Sparsh Sharma, Founder I am a product designer, passionate about building user focused fintech products. ### Company Launches [### Paasa - Build wealth across borders]( [Paasa]( gives high-net-worth individuals (HNIs) in emerging markets the same globally diversified portfolios as HNIs in America. **What’s the problem?** ‘Pre-HNIs’ and HNIs in countries like India want to diversify their wealth internationally. A few cross-border brokerage apps exist, but most customers hesitate due to: * Lack of guidance on foreign investments * Remittance regulations that change every year * Cross-border tax filing that their CFAs struggle with * Slow outflows from local banks These barriers keep global investing out of reach, reserved for the top 0.001% who can afford wealth managers and family offices. We’re changing that. Our vision is to bring all the opportunities of the 0.001% to the pre-HNI without any hassle. **Team** [Sparsh]( and I ([Nitish]( grew up in India and have known each other for 5 years. Both personal finance nerds, we previously worked at SoFi, where Sparsh led design on Student Loan Refinance, and I led engineering projects on Data Governance. **Ask** If you’re in India and want to invest globally, [reach out]( We’re [[email protected]](mailto:[email protected])

Phonely

phonely

https://phonely.ai

San Francisco, CA, USA

Phonely lets any business go from a website to a conversational AI receptionist in a matter of minutes. Our platform lets businesses connect to their scheduling software, route calls, or answer customer support questions from a knowledge base just like a human. After the call, Phonely provides powerful AI analytics and automatically extracts relevant information, sending it directly to the business CRM, phone, or inbox.

Go from a website to an AI receptionist in just a few minutes

B2B

B2B

S24

Active

Early

https://www.ycombinator.com/companies/phonely

https://yc-oss.github.io/api/batches/s24/phonely.json

Title: Phonely: Replicate your best human phone agents with voice AI | Y Combinator URL Source: Markdown Content: ### Replicate your best human phone agents with voice AI Phonely lets any business go from a website to a conversational AI receptionist in a matter of minutes. Our platform lets businesses connect to their scheduling software, route calls, or answer customer support questions from a knowledge base just like a human. After the call, Phonely provides powerful AI analytics and automatically extracts relevant information, sending it directly to the business CRM, phone, or inbox. Phonely Founded:2023 Team Size:4 Location:San Francisco ### Active Founders ### Will Bodewes, Founder Will Bodewes is the co-founder and CEO of Phonely. Prior to starting Phonely, he was pursuing a PhD at the University of Melbourne researching applied artificial intelligence applications. Before that, he founded an audio-tech startup that garnished celebrity engagement from Chance the Rapper, the band ASIA and more. During college, he competed in the NCAA's for XC skiing, and was one of the top ultra-endurance cyclists in the world. ### Nisal Ranasinghe, Founder Nisal is the Co-founder and CTO of Phonely. With a background in Electronic and Telecommunications Engineering, he worked as a senior engineer in the AI team of a major telecommunications company for nearly three years, where he led projects on advanced AI solutions. Nisal then pursued a PhD at the University of Melbourne, focusing on explainable AI and AI for healthcare. ### Company Launches [### 📞 Phonely - Humanlike AI receptionists]( **TL;DR We use your website to create an AI receptionist that can answer FAQs, route calls, and book appointments, ready to go in under 5 minutes. Don’t believe us? Try it free **[**here**]( **—** Hey 👋! It's [Will]( and [Nisal]( here. We’re building [Phonely]( an AI receptionist system that ensures **you’ll never get put on hold again.** We are both former PhD AI researchers from the University of Melbourne, and 8 months ago, we set out to build an AI receptionist that could do everything that a person could do over the phone, and it's safe to say we’re done just that. **❌ The Problem** Missed calls cost businesses tens of thousands of dollars in lost revenue every year. Sure, you can hire a receptionist or outsource to a call center, but even they face the same problem. Humans sleep, have bad days, and ultimately don’t scale. We can do it better, faster, and 70% cheaper than the status quo. **✨ Our Solution** Phonely uses a website URL to build a humanlike AI answering system in just a few minutes. Our platform then connects to scheduling software, routes calls, and integrates with an existing knowledge base, all while improving its responses over time. After the call, Phonely provides powerful AI analytics and automatically extracts relevant information, sending it directly to the business CRM, phone, or inbox. Phonely doesn’t just sound like a person. It can do things like a person, too. * Connect directly to your calendar to schedule, cancel, or rebook appointments * ⁠Send SMS messages * ⁠Trigger live automations based on the callers’ inquiry * Place outbound calls to remind customers about important events (i.e. upcoming appointments) * [And much more]( **⚙️ How it works** Phonely combines the best voice providers, language models, and transcription software to create contextually accurate, lifelike voice responses in under a second. If you give us your website, we will visit every page and extract relevant information automatically, formatting it to work with our custom-trained language models. We also directly integrate with a bunch of different software, such as Google Calendar and Zapier, to make it easy to manage appointments and so much more. See it in action: [ **🙏 Asks** 1. Build an AI receptionist from your business website (it's free and takes less than a minute). Do that [here]( 2. Don’t have a business but know one? We’re doing the best affiliate deal we’ll ever have. Join that [here]( 3. Also, if you have 10 minutes to chat about Phonely, please book a time with me [here]( 😊 Follow our [Linkedin]( [Twitter]( and [Youtube]( to stay up to date.

Autumn Labs

autumn-labs

https://www.autumnlabs.io/

San Francisco, CA, USA

Autumn Labs delivers a developer-friendly platform for monitoring and managing modern manufacturing lines. Seamlessly integrating with robotic stations—automated test stations, robotic assembly cells, and industrial arms—it ensures full traceability, live data monitoring, and secure data transport. We prioritize streamlining factory operations, boosting production quality, and preventing supply chain disruptions, all while offering an effortless onboarding experience for engineers and manufacturers.

A platform to monitor and manage industrial robots and test stations.

Industrials

Industrials -> Manufacturing and Robotics

S24

Active

Early

https://www.ycombinator.com/companies/autumn-labs

https://yc-oss.github.io/api/batches/s24/autumn-labs.json

Title: Autumn Labs: Datadog for industrial robots | Y Combinator URL Source: Markdown Content: ### Datadog for industrial robots Autumn Labs delivers a developer-friendly platform for monitoring and managing modern manufacturing lines. Seamlessly integrating with robotic stations—automated test stations, robotic assembly cells, and industrial arms—it ensures full traceability, live data monitoring, and secure data transport. We prioritize streamlining factory operations, boosting production quality, and preventing supply chain disruptions, all while offering an effortless onboarding experience for engineers and manufacturers. Autumn Labs Founded:2023 Team Size:3 Location:San Francisco ### Active Founders ### Prathik Muppidi, Founder University of Waterloo Mechatronics Engineering, ex-Google Engineer ### Harsh Bhatt, Founder Ex. Apple | Mechatronics Engineering @ University of Waterloo | Forever curious ### Sam Monga, Founder Hardware software engineer ### Company Launches [### Autumn Labs - Monitoring platform for industrial robots and manufacturing]( **tl;dr:** [Autumn Labs]( delivers a unified platform to monitor and manage modern factories. Our SDK seamlessly integrates with manufacturing systems (test stations, assembly robots, industrial arms, and manual stations) to provide real-time monitoring, control, and production data management. The result? Lower operation costs, higher product quality, and a smoother production line. 🍁 = 🤖 + 🏭 + 📊 ### **Problem - Advanced Machines, Outdated Infrastructure** The adoption of robots in manufacturing is exploding, especially to keep up with producing increasingly more complex products and shortages in skilled labor. However, the infrastructure in which these robots would be deployed is still years behind. Without effective systems, manufacturing physical products might seem like an insurmountable black box, especially when they are made in factories on another continent where critical information is only reported via spreadsheets every other week. The past: [_src_]( Today: [_src_]( But the manufacturing data is always lagging and still looks like this: ### **Solution - Seamless Integration, Real-time insights** Autumn Labs provides an SDK that’s easy to integrate into any robotic architecture, enabling seamless monitoring and management of machines and, ultimately, the factory in which they reside. This way, robotic integrators and developers have a standard platform where they pull and push critical data that’s specific to their machine’s tasks. Both factory operations and engineering teams are able to get actionable data they need to improve production utilizations and process quality. _A Unified Platform for Hybrid Factories:_ Our platform is built for hybrid factory lines, too, allowing for seamless integration of manual human assembly stations. Factory personnel can now get a holistic view of your factory's performance—robotic or not. With Autumn Labs, you can track, triage, and push fixes for: 1. Robotic assembly stations/cells 2. Automated test stations 3. Palletizing robots 4. Manual assembly stations ### **Traction - Real Results, Real Impact:** Autumn Labs is already monitoring and managing 50+ automated factory stations in 4 different consumer electronics factory lines across multiple companies and contract manufacturers. There’s a good chance that a product you use today has gone through our platform. Why our customers love us: 1. Clear insights into Manufacturing Value Added (MVA) costs 2. Increased machine utilization 3. Rapid improvements in production quality 4. Enhanced product development cycles ### **Meet the Founders** We’re [Sam]( [Harsh]( and [Pra]( — Mechatronics engineers from the University of Waterloo with decades of combined experience at Apple (iPhone), Google (Nest), and Meta (Oculus). After countless sleepless nights fixing robotic assembly lines and test stations, we decided it was time to create the ultimate tool for manufacturing. We’re building the solution we wish we had back then. ### **Our Ask - Let's Transform Manufacturing Together** Know someone struggling to get real-time data from robots, test stations, or factories? Let’s talk. Reach us at [[email protected]](mailto:[email protected])