Rag Agent Service

This project provides a secure Retrieval-Augmented Generation (RAG) agent service using FastAPI. The service allows you to store documents in a vector space and query them using LLM.

Features

• Store documents in a vector space (instead of in-memory vector store used session-based storage to save costs)
• Query the vector space using LLM
• Supports multiple document formats (PDF, etc.)

Requirements

• Python 3.8+
• pip

Installation

1. Clone the repository:

   git clone https://github.com/abhishekshingadiya/rag-agent.git
   cd secure-rag-agent

2. Install the required packages:

   pip install -r requirements.txt

3. Set up environment variables:

   cp env.example .env
   # Edit .env to add your OPENAI_API_KEY

Usage

1. Start the FastAPI server:

   uvicorn app:app --host 0.0.0.0 --port 8001

2. Access the health check endpoint:

   curl http://localhost:8001/health

3. Store documents in the vector space:

   curl -X POST "http://localhost:8001/secure-rag-agent/train" -H "Content-Type: application/json" -d '{"document_paths":["./data/handbook.pdf"]}'

4. Query the vector space(add session_id to reuse the vector space):

   curl -X POST "http://localhost:8001/secure-rag-agent/query" -H "Content-Type: application/json" -d '{"questions": ["What is the name of the company?", "Who is PM of India?", "Who is the CEO of the company?", "What is their vacation policy?", "What is the termination policy?"], "session_id": "your_session_id"}'

5. Store documents and query the vector space:

   curl -X POST 'http://localhost:8001/secure-rag-agent' -H 'accept: application/json' -H 'Content-Type: application/json' -d '{"document_paths":["./data/handbook.pdf"], "questions":["What is the name of the company?", "Who is PM of India?", "Who is the CEO of the company?", "What is their vacation policy?", "What is the termination policy?"]}'

Endpoints

• GET /health: Check if the service is running.
• POST /secure-rag-agent: Store documents and query the vector space.
• POST /secure-rag-agent/train: Store documents in the vector space.
• POST /secure-rag-agent/query: Query the vector space using LLM.

License

This project is licensed under the MIT License.