Insurance Policy Analysis using LLMs

A powerful document analysis application that leverages Large Language Models (LLMs) to provide intelligent question-answering capabilities for insurance policies and other documents. The application uses RAG (Retrieval-Augmented Generation) to provide accurate, context-aware responses based on uploaded documents.

Features

• Multi-LLM Support: Choose between OpenAI, Azure OpenAI, and Google Gemini models
• Document Upload: Support for PDF and DOCX files
• Vector Search: FAISS-based vector database for efficient document retrieval
• Interactive Chat: Streamlit-powered web interface with chat history
• RAG Implementation: Combines document retrieval with LLM generation for accurate responses
• Persistent Storage: Embeddings are stored and can be reused across sessions

Technology Stack

• Frontend: Streamlit
• LLMs: OpenAI GPT-4o, Azure OpenAI, Google Gemini
• Vector Database: FAISS
• Document Processing: LangChain, PyPDF, Unstructured
• Embeddings: OpenAI text-embedding-3-large, Azure OpenAI, Google Generative AI

Prerequisites

Before running the application, you'll need API keys for at least one of the following services:

• OpenAI API Key (for OpenAI models)
• Azure OpenAI credentials (for Azure OpenAI models)
• Google API Key (for Gemini models)

Quick Start

Local Development

1. Clone the repository:

   git clone https://github.com/kalpshah18/InsurancePolicyAnalysis.git
   cd InsurancePolicyAnalysis/backend

2. Install dependencies:

   pip install -r requirements.txt

3. Set up environment variables:

   cp .env.example .env
   # Edit .env with your actual API keys

4. Run the application:

   streamlit run streamlit_app.py

Streamlit Cloud Deployment

For detailed deployment instructions, see DEPLOYMENT.md.

Quick Deploy:

1. Fork this repository on GitHub
2. Go to share.streamlit.io
3. Create a new app from your forked repository
4. Add your API keys in the Secrets section
5. Your app will be live at https://[your-app-name].streamlit.app

Configuration

You'll need API keys for at least one of the following services:

Azure OpenAI Configuration (if using Azure)

API_KEY=your_azure_openai_api_key API_BASE=your_azure_openai_endpoint EMBEDDING_DEPLOYMENT_NAME=your_embedding_deployment_name AZURE_AI_SEARCH_API_VERSION=2023-05-15

Google Gemini Configuration (if using Gemini)

GOOGLE_API_KEY=your_google_api_key

## Usage

1. **Start the application**:
```bash
streamlit run streamlit_app.py

2. Access the web interface: Open your browser and navigate to http://localhost:8501

3. Using the application:

   • Select your preferred LLM model from the sidebar (OpenAI, Azure OpenAI, or Gemini)
   • Upload a document (PDF or DOCX) using the file uploader
   • Click "Submit" to process the document and create embeddings
   • Start asking questions about your document in the chat interface

Project Structure

backend/
├── src/
│   ├── __init__.py
│   ├── document_loader.py      # Document loading and processing
│   ├── embed_and_store.py      # Vector embeddings and FAISS storage
│   ├── llm_model.py           # LLM model configurations
│   └── retrieval_qa.py        # RAG implementation
├── faiss_index/               # Vector database storage (Created Automatically when Program is Runw)
├── streamlit_app.py          # Main Streamlit application
├── requirements.txt          # Python dependencies
├── README.md                # Project documentation

How It Works

1. Document Processing: Uploaded documents are processed using LangChain loaders (PyPDF for PDFs, Unstructured for DOCX)
2. Embedding Generation: Text chunks are converted to vector embeddings using the selected embedding model
3. Vector Storage: Embeddings are stored in a FAISS vector database for efficient similarity search
4. Query Processing: User questions are embedded and used to retrieve relevant document chunks
5. Response Generation: Retrieved context is combined with the user question and sent to the LLM for answer generation

Use Cases

• Insurance Policy Analysis: Analyze complex insurance documents and get quick answers
• Contract Review: Upload contracts and ask specific questions about terms and conditions
• Document Q&A: General purpose document question-answering for various file types
• Research Assistant: Extract specific information from large documents efficiently

Configuration

LLM Models

• OpenAI: Uses GPT-4o by default with temperature 0.2
• Azure OpenAI: Configurable deployment and API version
• Gemini: Google's Generative AI models

Embedding Models

• OpenAI: text-embedding-3-large
• Azure OpenAI: Configurable embedding deployment
• Gemini: Google Generative AI embeddings

Security Notes

• Store API keys securely in the .env file
• Never commit API keys to version control
• The .env file should be added to .gitignore

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Troubleshooting

• "Please upload a document first": Make sure to upload a document and click "Submit" before asking questions
• API Key errors: Verify your API keys are correctly set in the .env file
• Import errors: Ensure all dependencies are installed with pip install -r requirements.txt

Support

For support, please open an issue at kalpshah18/InsurancePolicyAnalysis or contact @kalpshah18 on GitHub.