Lang-RAG

A production-ready Retrieval-Augmented Generation (RAG) system built with FastAPI, React, ChromaDB, and Hugging Face embeddings. Upload PDF documents, index them into a vector database, and interact with them through a conversational AI interface with source-grounded responses.

---

Features

Document Management

• Upload single or multiple PDF documents
• Automatic PDF text extraction
• Document listing and deletion
• Persistent document storage

RAG Pipeline

• PDF parsing and text cleaning
• Intelligent text chunking with overlap
• Hugging Face sentence-transformer embeddings
• ChromaDB vector storage and retrieval
• Metadata-aware document indexing

Conversational AI

• Context-aware question answering
• Source-grounded responses
• Conversation history support
• Streaming responses using Server-Sent Events (SSE)
• Confidence-based document retrieval

Modern Frontend

• React + Vite
• Responsive UI
• Drag-and-drop file uploads
• Real-time chat interface
• Markdown rendering
• Syntax highlighting
• Upload progress tracking

Deployment Ready

• FastAPI backend
• Vercel frontend deployment
• Render backend deployment
• Docker support
• Environment-based configuration

---

Tech Stack

Backend

• FastAPI
• LangChain
• ChromaDB
• Sentence Transformers
• Hugging Face Embeddings
• PyPDF
• Uvicorn

Frontend

• React
• Vite
• Axios
• Tailwind CSS
• Framer Motion

Vector Database

• ChromaDB

---

Project Structure

lang-rag/
│
├── backend/
│   ├── app/
│   │   ├── api/
│   │   ├── core/
│   │   ├── models/
│   │   ├── services/
│   │   └── utils/
│   │
│   ├── uploads/
│   ├── vectorstore/
│   ├── requirements.txt
│   └── main.py
│
├── frontend/
│   ├── src/
│   ├── public/
│   ├── package.json
│   └── vite.config.js
│
├── docker-compose.yml
├── .env.example
└── README.md

---

Installation

Clone Repository

git clone https://github.com/Kruthica/lang-rag.git
cd lang-rag

---

Backend Setup

cd backend

python -m venv .venv

Activate Environment

Windows:

.venv\Scripts\activate

Linux/macOS:

source .venv/bin/activate

Install Dependencies

pip install -r requirements.txt

Configure Environment

Create a .env file inside the backend folder:

HF_TOKEN=your_huggingface_token

UPLOAD_DIR=uploads
CHROMA_DB_DIR=vectorstore

CORS_ORIGINS=http://localhost:5173,http://localhost:5174

Run Backend

uvicorn main:app --reload --host 0.0.0.0 --port 8000

Backend API:

http://localhost:8000

Swagger Documentation:

http://localhost:8000/docs

---

Frontend Setup

cd frontend

npm install
npm run dev

Frontend:

http://localhost:5173

---

API Endpoints

Health Check

GET /health

Upload Documents

POST /upload

Ask Questions

POST /ask

Streaming Answers

POST /ask/stream

List Documents

GET /documents

Delete Document

DELETE /documents/{document_id}

---

Example Requests

Upload PDF

curl -X POST http://localhost:8000/upload \
-F "files=@sample.pdf"

Ask Question

curl -X POST http://localhost:8000/ask \
-H "Content-Type: application/json" \
-d '{
  "question":"Summarize this document",
  "history":[]
}'

Get Documents

curl http://localhost:8000/documents

---

Deployment

Frontend (Vercel)

Set:

VITE_API_URL=https://your-backend-url.onrender.com

Deploy:

vercel --prod

Backend (Render)

Set environment variables:

HF_TOKEN=your_token

Start command:

uvicorn main:app --host 0.0.0.0 --port $PORT

---

Future Improvements

• Hybrid search (vector + keyword)
• Reranking pipelines
• OCR support for scanned PDFs
• Multi-user authentication
• Role-based document access
• Citation highlighting
• Document collections and folders
• Voice-based interactions

---

License

MIT License

---

Author

Kruthica

Built as a production-focused Retrieval-Augmented Generation system for document intelligence and conversational knowledge retrieval.