DRAG, a Document Search API

A lightweight, cost-effective Retrieval-Augmented Generation (RAG) API built for efficient document search and context-aware LLM responses.

Tech Stack

• Framework: FastAPI
• Orchestration: LlamaIndex
• Database: PostgreSQL with pgvector extension
• Dependency Management: uv
• Hosting: Render

Architecture

Local Development & Setup

1. Clone the repository:

   git clone https://github.com/your-username/your-repo.git
   cd your-repo

2. Install dependencies using uv:

   uv sync

3. Configure environment variables: Create a .env file in the root directory and add your credentials:

   OPEN_API_KEY=your_open_api_key_here
   DATABASE_URL=postgresql://user:password@localhost:5432/dbname
   MODEL=model_name

4. Add custom documents to data/raw_docs directory (optional)

5. Run the development server:

   uv run