AI-Powered Multilingual Government Scheme Intelligence System
Bharat Sahayak AI is an intelligent platform that helps Indian citizens discover government schemes they're eligible for. Using natural language processing and AI reasoning, it simplifies the complex landscape of 1000+ government schemes into personalized, actionable recommendations.
- Awareness Gap: Citizens don't know which schemes apply to them
- Language Barrier: Schemes explained in English/complex language
- Eligibility Complexity: Multi-factor criteria are hard to understand
- Middlemen Dependency: Leads to fraud and misinformation
- Low Utilization: Eligible citizens never apply
- ✅ Natural language input in multiple Indian languages
- ✅ AI-powered eligibility reasoning (explains WHY you qualify)
- ✅ Multilingual explanations in simple language
- ✅ Voice interface for semi-literate users
- ✅ Document intelligence for PDF processing
- ✅ Step-by-step application guidance
Frontend:
- HTML5, CSS3, JavaScript
- Responsive design for mobile and desktop
Backend:
- Python 3.11
- Flask web framework
- RESTful API
AI/ML Services (AWS):
- Amazon Bedrock (Claude 3 Sonnet) - Core reasoning and NLP
- Amazon Comprehend - Entity extraction and language detection
- Amazon Textract - PDF text extraction
- Amazon Translate - Multi-language translation
- Amazon Polly - Text-to-speech
- Amazon Transcribe - Speech-to-text
Infrastructure (AWS):
- AWS Lambda - Serverless compute
- Amazon DynamoDB - Scheme database
- Amazon S3 - Document storage
- Amazon API Gateway - API management
- AWS CloudWatch - Monitoring and logging
User Input (Text/Voice)
↓
API Gateway
↓
Lambda: Profile Extractor (Comprehend + Bedrock)
↓
Structured User Profile
↓
Lambda: Scheme Matcher (DynamoDB + Bedrock)
↓
Matched Schemes with Reasoning
↓
Lambda: Translator (Amazon Translate)
↓
Regional Language Output
↓
Lambda: Voice Output (Amazon Polly - if requested)
↓
User Interface
- Python 3.11 or higher
- pip (Python package manager)
- Git
- Clone the repository
git clone https://github.com/yourusername/bharat-sahayak-ai.git
cd bharat-sahayak-ai- Install dependencies
pip install -r requirements.txt- Run the application
python app.py- Access the web interface
Open your browser and navigate to:
http://localhost:5000
Example 1: Farmer in Maharashtra
Input: "I am a farmer from Maharashtra, 45 years old, my annual income is 1.8 lakh, I own 2 acres of farmland"
Output:
✅ PM-KISAN - ₹6,000/year
✅ Maharashtra Mukhyamantri Kisan Yojana - ₹6,000/year
Example 2: Senior Citizen
Input: "I am 65 years old, my income is below 1 lakh per year, I have a BPL card, from Delhi"
Output:
✅ Indira Gandhi National Old Age Pension Scheme - ₹200-500/month
✅ PM Ujjwala Yojana - Free LPG connection
Example 3: Girl Child Education
Input: "I have a 5-year-old daughter, I want to save for her education"
Output:
✅ Sukanya Samriddhi Yojana - 8.2% interest with tax benefits
-
Intelligent Profile Extraction
- Understands natural language input
- Mixed language support (e.g., Hindi-English)
- Extracts: age, income, occupation, location, category
-
AI-Powered Eligibility Reasoning
- Not just matching, but explaining WHY
- Example: "You qualify because: farmer + land ownership + income < ₹2L"
- Relevance scoring (0-100)
-
Multilingual Support
- Input and output in 10+ Indian languages
- Simple, conversational explanations
- Culturally appropriate context
-
Document Intelligence
- Processes government PDF circulars
- Extracts eligibility criteria automatically
- Summarizes in simple language
-
Voice Interface (Coming Soon)
- Voice input for semi-literate users
- Voice output in regional languages
- Critical for rural penetration
-
Personalized Guidance
- Required documents checklist
- Step-by-step application process
- Nearest help center information
- Helpline numbers
Current coverage:
- 7+ Central government schemes
- State-specific schemes (Maharashtra as pilot)
- Categories: Agriculture, Housing, Social Security, Education, Energy, MSME
Schemes included:
- PM-KISAN
- PMAY (Pradhan Mantri Awas Yojana)
- Indira Gandhi National Old Age Pension
- Sukanya Samriddhi Yojana
- PM Ujjwala Yojana
- PMEGP (Employment Generation Programme)
- Maharashtra Mukhyamantri Kisan Yojana
http://localhost:5000/api
POST /api/analyze
Content-Type: application/json
{
"text": "I am a farmer from Maharashtra, 45 years old, income 1.8 lakh",
"language": "en"
}Response:
{
"success": true,
"profile": {
"age": 45,
"income": 180000,
"occupation": "Farmer",
"state": "Maharashtra",
"hasLand": true
},
"totalMatches": 2,
"schemes": [
{
"schemeId": "PMKISAN-001",
"name": "PM-KISAN",
"category": "Agriculture",
"benefits": "₹6,000 per year",
"relevanceScore": 85,
"explanation": "You are eligible for PM-KISAN...",
"documents": ["Aadhaar", "Land records", "Bank account"],
"applicationProcess": "Apply at pmkisan.gov.in",
"helpline": "155261"
}
],
"timestamp": "2024-01-15T10:30:00"
}GET /api/schemesGET /api/scheme/{scheme_id}GET /api/health- Color Palette: Teal Trust (aligned with AWS branding)
- Responsive Design: Works on mobile, tablet, and desktop
- Accessibility: High contrast, clear typography
- Language Toggle: Easy switching between languages
- Example Prompts: Quick-start templates for users
- Flask development server
- In-memory scheme database
- Single instance
Phase 1 (Month 1-2):
- Migrate to AWS Lambda + API Gateway
- DynamoDB for scheme database
- Amazon Bedrock integration for real AI reasoning
- 100 central schemes coverage
Phase 2 (Month 3-4):
- Add 500+ state schemes
- 10 language support via Amazon Translate
- Voice interface via Polly + Transcribe
- Mobile app (React Native)
Phase 3 (Month 5-6):
- DigiLocker integration
- Direct application submission
- CSC (Common Service Center) deployment
- 10 state government partnerships
Target Metrics:
- 100,000 users by Month 3
- 70% satisfaction score
- 50% reduction in agent dependency
- No PII Storage: User profiles auto-deleted after 24 hours
- Encryption: TLS 1.3 in transit, AWS KMS at rest
- Compliance: Aligned with India's Digital Personal Data Protection Act
- Rate Limiting: Prevents abuse and DDoS
- Data Residency: AWS Mumbai region
We welcome contributions! Please follow these steps:
- Fork the repository
- Create a feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
# Create virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
pip install -r requirements.txt
# Run tests
python -m pytest tests/
# Run locally
python app.pyThis project is licensed under the MIT License - see the LICENSE file for details.
AI for Bharat Hackathon 2024-25
- Your Name - Project Lead & Backend Developer
- Team Member 2 - Frontend Developer
- Team Member 3 - AI/ML Engineer
- Team Member 4 - Data Engineer
- AWS for cloud infrastructure and AI services
- YourStory for organizing AI for Bharat hackathon
- Government of India for open data on schemes
- Community for feedback and testing
- Email: contact@bharatsahayak.ai
- GitHub: https://github.com/yourusername/bharat-sahayak-ai
- Presentation: [Link to presentation deck]
Our vision is to create an inclusive AI platform that:
- Empowers every Indian citizen with scheme awareness
- Reduces dependency on middlemen and agents
- Increases government scheme utilization by 50%
- Supports 100+ million citizens in discovering benefits
Empowering Every Indian Citizen with AI 🇮🇳
Built with ❤️ for Bharat