Asaad Almutareb
fixed the knowledgeBase_search function to use the sources cache properly
fba76a5
|
raw
history blame
1.64 kB
metadata
title: Insurance Advisor Agents PoC
emoji: 🤖
colorFrom: red
colorTo: indigo
sdk: gradio
python: 3.11
app_file: app_gui.py
pinned: false

Insurance Advisor Agent(s)

Welcome to the Insurance Advisor Agent(s) project! This project aims to set up a modular, multi-agent system to handle inquiries for an insurance company. The system utilizes various approaches to provide reliable answers regarding insurance products.

Features

Vectorstore Search

  • Multi-index (chunks + summaries)
  • Metadata Filtering
  • Re-ranking [DONE]
  • Hybrid Search (BM25 + Vectorstore) [DONE]
  • HyDE/HyQE

ReAct Agent

  • Usage of ReAct Agent instead of Chain [DONE]
  • Added Tools (Google Search Engine) [DONE]
  • Query Re-writing
  • Additional Enhancements

Corrective RAG

  • Implementation of Corrective Retrieval-Augmented Generation

Workflow for Agents

  • Determination of the customer's funnel stage
  • Custom prompts for each step
  • Following a structured "script"

Getting Started

To get started with this project, follow these steps:

  1. Clone the Repository:

    git clone https://github.com/your-username/insurance-advisor-agent.git
    cd insurance-advisor-agent
    
  2. Install Dependencies:

    pip install -r requirements.txt
    
  3. Run the Application:

    python main.py
    

Contributing

We welcome contributions! Please read our Contributing Guide to get started. License

This project is licensed under the MIT License. See the LICENSE file for details. Contact

If you have any questions or suggestions, feel free to open an issue or start a discussion.