Skip to content
BranchesTags

Repository files navigation

SevaSetu AI 🇮🇳

AI-Powered Public Service Assistant for Government Scheme Guidance

"Bridging Citizens to Government Services" Developed by Rahul Jha | Made in India 🇮🇳

Python FastAPI React MySQL Docker License

🌟 Features

Feature Technology
🤖 AI Chat (RAG) Gemini API + ChromaDB + Sentence-BERT
🗳️ 10+ Government Services Voter ID, PAN, Passport, Certificates
🏛️ 100+ Government Schemes Central + State, Eligibility Scoring
📄 Document OCR Tesseract (EN+HI+MR), Field Extraction
🌐 Multilingual English, Hindi, Marathi
🎤 Voice Input/Output Web Speech API
🔐 Auth JWT + bcrypt + RBAC
📊 Admin Panel Analytics, User & Scheme Management
📥 PDF Reports ReportLab — Professional PDFs
🐳 Deployment Docker Compose + Nginx

🏗️ Project Structure

sevasetu-ai/
├── frontend/                    # React.js Frontend
│   ├── src/
│   │   ├── App.jsx              # Root app + auth context + routing
│   │   ├── App.css              # Global design system CSS
│   │   ├── pages/
│   │   │   ├── SplashScreen.jsx # 3D animated intro screen
│   │   │   ├── LoginPage.jsx    # JWT login
│   │   │   ├── RegisterPage.jsx # User registration
│   │   │   ├── Dashboard.jsx    # User dashboard with stats
│   │   │   ├── ChatPage.jsx     # AI chatbot interface ⭐
│   │   │   ├── SchemesPage.jsx  # Scheme browser + eligibility
│   │   │   ├── DocumentsPage.jsx# Document upload + OCR + checklist
│   │   │   ├── HistoryPage.jsx  # Query history
│   │   │   ├── ReportsPage.jsx  # PDF report generation
│   │   │   ├── AdminPage.jsx    # Admin panel
│   │   │   └── ProfilePage.jsx  # User profile settings
│   │   ├── components/
│   │   │   ├── MainLayout.jsx   # Sidebar + topnav layout
│   │   │   ├── LoadingSpinner.jsx
│   │   │   └── ...
│   │   └── services/
│   │       └── api.js           # Axios API client (all endpoints)
│   ├── public/
│   ├── package.json
│   └── Dockerfile
│
├── backend/                     # Python FastAPI Backend
│   ├── app/
│   │   ├── main.py              # FastAPI app entry point ⭐
│   │   ├── core/
│   │   │   ├── config.py        # Pydantic settings
│   │   │   ├── auth.py          # JWT + bcrypt + RBAC
│   │   │   ├── database.py      # Async SQLAlchemy + MySQL
│   │   │   └── logging_config.py
│   │   ├── api/
│   │   │   ├── auth.py          # /auth/* endpoints
│   │   │   ├── queries.py       # /queries/* (RAG pipeline) ⭐
│   │   │   ├── schemes.py       # /schemes/* CRUD + eligibility
│   │   │   ├── documents.py     # /documents/* upload + OCR
│   │   │   ├── reports.py       # /reports/* PDF generation
│   │   │   ├── admin.py         # /admin/* analytics
│   │   │   ├── users.py         # /users/* profile
│   │   │   └── __init__.py
│   │   └── models/
│   │       └── user.py          # All SQLAlchemy models
│   ├── requirements.txt
│   └── Dockerfile
│
├── ai/                          # AI/ML Modules
│   ├── rag/
│   │   └── rag_engine.py        # RAG pipeline (ChromaDB + Gemini) ⭐
│   ├── ocr/
│   │   └── ocr_engine.py        # Tesseract OCR + field extraction ⭐
│   └── embeddings/
│       └── embedding_service.py
│
├── database/
│   └── schema.sql               # MySQL schema + seed data ⭐
│
├── docs/
│   └── PROJECT_DOCUMENTATION.md # Complete SRS, Literature Survey, Viva QA ⭐
│
├── docker/
│   ├── docker-compose.yml       # Full stack deployment ⭐
│   └── nginx/
│       └── nginx.conf
│
└── README.md                    # This file

🚀 Quick Start

Prerequisites

  • Python 3.11+
  • Node.js 18+
  • MySQL 8.0
  • Tesseract OCR (with Hindi+Marathi packs)
  • Docker & Docker Compose (for containerized deployment)

1. Clone Repository

git clone https://github.com/rahul-jha/sevasetu-ai.git
cd sevasetu-ai

2. Configure Environment

# Backend .env
cp backend/.env.example backend/.env

# Edit backend/.env:
SECRET_KEY=your-super-secret-key-here
DATABASE_URL=mysql+asyncmy://sevasetu:password@localhost:3306/sevasetu_db
GEMINI_API_KEY=your-gemini-api-key-here    # Set this only in your backend host environment; never commit the real key    # Get from aistudio.google.com
REDIS_URL=redis://localhost:6379

# Frontend .env
echo "REACT_APP_API_URL=http://localhost:8000" > frontend/.env

3. Database Setup

mysql -u root -p < database/schema.sql

4. Backend Setup

cd backend
python -m venv venv
source venv/bin/activate          # Windows: venv\Scripts\activate
pip install -r requirements.txt

# Install Tesseract OCR (Ubuntu/Debian)
sudo apt-get install -y tesseract-ocr tesseract-ocr-hin tesseract-ocr-mar

# Start backend
uvicorn app.main:app --reload --port 8000

5. Frontend Setup

cd frontend
npm install
npm start
# Opens at http://localhost:3000

6. Access Application

Service URL
🌐 Frontend http://localhost:3000
🔌 Backend API http://localhost:8000
📖 Swagger Docs http://localhost:8000/docs
📊 ReDoc http://localhost:8000/redoc

Default Admin Login:

  • Email: admin@sevasetu.ai
  • Password: Admin@SevaSetu1

🐳 Docker Deployment

# Full stack with Docker Compose
cd docker
cp .env.example .env              # Configure your API keys
docker-compose up --build         # Development
docker-compose up -d              # Production (detached)

# Check logs
docker-compose logs -f backend
docker-compose logs -f frontend

# Stop
docker-compose down

☁️ Cloud Deployment (AWS/GCP)

AWS Elastic Beanstalk

# Install EB CLI
pip install awsebcli

# Initialize
eb init sevasetu-ai --region ap-south-1 --platform docker

# Create environment
eb create production --instance_type t3.medium

# Deploy
eb deploy

Google Cloud Run

# Build and push images
gcloud builds submit --tag gcr.io/PROJECT_ID/sevasetu-backend ./backend
gcloud builds submit --tag gcr.io/PROJECT_ID/sevasetu-frontend ./frontend

# Deploy backend
gcloud run deploy sevasetu-backend \
  --image gcr.io/PROJECT_ID/sevasetu-backend \
  --region asia-south1 \
  --allow-unauthenticated \
  --set-env-vars GEMINI_API_KEY=xxx

# Deploy frontend
gcloud run deploy sevasetu-frontend \
  --image gcr.io/PROJECT_ID/sevasetu-frontend \
  --region asia-south1 \
  --allow-unauthenticated

🔌 API Quick Reference

# Register
curl -X POST http://localhost:8000/api/v1/auth/register \
  -H "Content-Type: application/json" \
  -d '{"name":"Test","email":"test@example.com","mobile":"9876543210","password":"Test@1234","state":"Maharashtra"}'

# Login
curl -X POST http://localhost:8000/api/v1/auth/login \
  -d "username=test@example.com&password=Test@1234"

# Ask AI (RAG)
curl -X POST http://localhost:8000/api/v1/queries/ask \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"question":"How to apply for PAN card?","language":"en"}'

# Check scheme eligibility
curl -X POST http://localhost:8000/api/v1/schemes/eligibility \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"annual_income":200000,"age":35,"gender":"male","is_farmer":true}'

# Upload document for OCR
curl -X POST http://localhost:8000/api/v1/documents/upload \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -F "file=@aadhaar.jpg" \
  -F "service_type=voter_id"

🧪 Running Tests

cd backend
pytest tests/ -v --asyncio-mode=auto

# With coverage report
pytest tests/ --cov=app --cov-report=html
open htmlcov/index.html

🤖 Getting Gemini API Key

  1. Go to Google AI Studio
  2. Sign in with your Google account
  3. Click "Get API Key" → "Create API Key"
  4. Copy key into backend/.env as GEMINI_API_KEY=xxx

Free tier: 15 requests/minute, 1 million tokens/month — sufficient for development and testing.

📊 Tech Stack Summary

Layer Technology Purpose
Frontend React.js 18, React Router 6 Single Page Application
Styling CSS Modules, Custom Properties Design system
API Client Axios + interceptors HTTP + auth
Backend FastAPI + Uvicorn REST API
Auth JWT + bcrypt Authentication + RBAC
Database MySQL 8 + SQLAlchemy async Relational data
Cache Redis 7 Sessions + rate limit
AI/LLM Google Gemini 1.5 Flash Language generation
Vector DB ChromaDB Semantic search
Embeddings Sentence-BERT (MiniLM) Query embedding
OCR Tesseract 5 + OpenCV Document processing
PDF ReportLab Report generation
Container Docker + Docker Compose Deployment
Web Server Nginx Reverse proxy + SSL

👨‍💻 Author

Rahul Jha

  • Final Year AIML Student
  • Project: SevaSetu AI — Government Services Assistant
  • Made with ❤️ in India 🇮🇳
  • Contact: rahul@sevasetu.ai

📜 License

MIT License — Free to use for educational purposes.

Copyright (c) 2026 Rahul Jha

🙏 Acknowledgements

  • Google Gemini team for the powerful LLM API
  • ChromaDB team for the vector database
  • Hugging Face for Sentence-Transformers
  • FastAPI team for the excellent Python web framework
  • Government of India for the open scheme data

Jai Hind! 🇮🇳 — SevaSetu AI: Bridging 1.4 Billion Citizens to Government Services

Render deployment

This project is configured for Render with render.yaml. Deploy the repository as a Blueprint. It creates a FastAPI Docker Web Service and a React Static Site.

Set these backend secrets in Render: GEMINI_API_KEY, DATABASE_URL, REDIS_URL, SMTP_USER, and SMTP_PASSWORD. The backend also generates a production SECRET_KEY.

The frontend uses REACT_APP_API_URL and must point to the Render FastAPI service root URL (do not append /api/v1).

For Forgot Password, configure Gmail SMTP with a Gmail App Password so the backend can send reset links.

For the complete Render deployment process, see RENDER_DEPLOYMENT.md.

About

No description, website, or topics provided.

Resources

Readme
MIT license
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

Packages

Contributors

Languages