DocMind — RAG Chat Application

A full-stack Retrieval-Augmented Generation (RAG) application that lets users upload PDF documents and have intelligent conversations about them, powered by Google Gemini and FastAPI.

---

Features

• PDF Upload & Processing — Upload any PDF and it gets converted into vector embeddings via FAISS
• Conversational AI — Chat with your document using Google Gemini 2.5 Flash
• Full Auth System — JWT-based authentication with signup and login
• Persistent Chat History — All conversations saved to PostgreSQL, resumable anytime
• ChatGPT-like Sidebar — Switch between multiple document chats
• React Frontend — Clean, dark-themed UI built with Vite + React

---

Tech Stack

Layer	Technology
Frontend	React 18, Vite
Backend	FastAPI, Python 3.12
AI Model	Google Gemini 2.5 Flash
Vector Store	FAISS
Database	PostgreSQL 16
ORM	SQLAlchemy + Alembic
Auth	JWT (PyJWT) + bcrypt

---

Project Structure

docmind/
├── app.py                  # FastAPI main application
├── config.py               # Environment config
├── requirements.txt        # Python dependencies
│
├── api/
│   ├── auth.py             # Signup & login routes
│   └── security.py         # JWT token validation
│
├── dbsetup/
│   ├── db_config.py        # Database connection
│   ├── models.py           # SQLAlchemy models
│   └── migrations/         # Alembic migrations
│
├── dependencies/
│   ├── prepare_pdf.py      # PDF → FAISS vector index
│   └── prepare_prompt.py   # RAG prompt builder
│
└── frontend/               # React application
    └── src/
        └── App.jsx         # Main UI component

---

Getting Started

Prerequisites

• Python 3.12
• Node.js 20+
• PostgreSQL 16
• Google Gemini API key — get one at aistudio.google.com

---

1. Clone the Repository

git clone https://github.com/YOUR_USERNAME/docmind.git
cd docmind

2. Set Up Python Environment

python -m venv gemni_env
source gemni_env/bin/activate
pip install -r requirements.txt

3. Set Up PostgreSQL

sudo -u postgres psql

CREATE DATABASE rag_app;
CREATE USER your_user WITH PASSWORD 'your_password';
GRANT ALL PRIVILEGES ON DATABASE rag_app TO your_user;
\q

4. Configure Environment Variables

Create a .env file in the root directory:

DATABASE_URL=postgresql://your_user:your_password@localhost/rag_app
SECRET_KEY=your_super_secret_key_here
ALGORITHEM=HS256
API_KEY=your_google_gemini_api_key
PORT=8000

5. Run Database Migrations

alembic upgrade head

6. Start the Backend

python app.py

Backend runs at http://localhost:8000 Swagger docs at http://localhost:8000/docs

7. Set Up and Start the Frontend

cd frontend
npm install
npm run dev

Frontend runs at http://localhost:5173

---

API Endpoints

Auth

Method	Endpoint	Description
POST	/api/v1/user/auth/signup	Register new user
POST	/api/v1/user/auth/login	Login, returns JWT token

Chat

Method	Endpoint	Description
POST	/api/upload	Upload a PDF, creates a chat session
POST	/api/chat	Send a message in a session
GET	/api/chat/sessions	Get all sessions for the user
GET	/api/chat/{session_id}/messages	Get full message history

---

How It Works

User uploads PDF
      ↓
PDF split into chunks → embedded into vectors → saved as FAISS index (named by session_id)
      ↓
User sends a message
      ↓
Relevant chunks retrieved from FAISS → injected into Gemini prompt
      ↓
Gemini responds using document context
      ↓
Both messages saved to PostgreSQL
      ↓
User can resume any past conversation from the sidebar

---

Environment Variables Reference

Variable	Description
DATABASE_URL	PostgreSQL connection string
SECRET_KEY	Secret key for signing JWT tokens
ALGORITHEM	JWT algorithm (e.g. HS256)
API_KEY	Google Gemini API key
PORT	Port for uvicorn (default 8000)

---

License

MIT License — feel free to use and modify.