🚀 FastAPI Document Management Backend

Overview

This is a modular, production-ready backend built with FastAPI for managing users, documents, and AI-assisted semantic search. The project leverages PostgreSQL + pgvector, SQLAlchemy, Alembic, and integrates LangChain + OpenAI (GPT-4o) for powerful document-based LLM workflows.

Core Features

• User Registration & Authentication:
  • Register, login, and secure endpoints via token-based access control.
• Document Uploads & Management:
  • Upload and manage documents via API.
  • Each document is embedded and stored as a vector for semantic retrieval.
• Semantic Document Search:
  • Vector-based document search using pgvector.
  • Users can query for documents by meaning, not just keywords, via endpoints.
• LangChain & AI Agent Integration:
  • The codebase is already integrated with LangChain and supports LLM-based workflows (e.g., OpenAI GPT-4o).
  • Tools for natural language queries, summarization, and contextual search.
  • Core logic in app/agent/rag_agent.py and rag_helper.py.

Roadmap/Next Steps

• DevContainer Support:
  • Add .devcontainer/ with PostgreSQL & Python setup for VS Code.
• Pluggable AI Tooling:
  • Extend get_tools() to include OCR, summarization, translation, etc.
• User Roles & Permissions:
  • Add roles (admin, editor, viewer) with per-document permissions.
• Frontend Development:
  • Build a basic React/Next.js frontend for UX.
• Performance Optimization & Deployment:
  • Containerize the app with Docker, add production-ready settings, and prepare for cloud deployment (e.g., Railway, Azure, AWS).

⚙️ Installation

Prerequisites

Ensure you have the following installed:

• Python 3.12+
• PostgreSQL
• (Optional) virtualenv or venv

Setup

1. Clone the repository:

   git clone <repository-url>
   cd fastapi_project

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows use `.venv\Scripts\activate`

3. Install dependencies:

   pip install -r requirements.txt

4. Set up environment variables (e.g., .env file for database connection):

   DATABASE_URL=postgresql://user:password@localhost/dbname
   OPENAI_API_KEY=your-open-ai-key
   

5. Run database migrations:

   alembic revision --autogenerate -m "migration commit message"
   alembic upgrade head

Running the Application

To start the FastAPI server, use:

uvicorn main:app --reload

📝 Notes

• Project uses pgvector for native similarity search in PostgreSQL.
• Designed for modular LLM agent workflows using LangChain + OpenAI.
• Core logic and agent tools can be extended easily for new AI capabilities.