आरोग्य सेतु (ArogyaSetu) - AI-Powered Healthcare Companion 🏥

Webiste Link

https://arogya-setu-client.vercel.app/

🏆 DUHacks 4.0 Hackthon- Best Healthcare Innovation

"Transforming healthcare access through WhatsApp - because health guidance should be just a message away."

💡 The Problem We're Solving

In India, millions lack access to basic healthcare guidance. Many hesitate to seek medical help due to:

• Cost barriers
• Limited access to doctors
• Language barriers
• Time constraints
• Fear of hospitals

🌟 Our Solution

आरोग्य सेतु (ArogyaSetu) is a revolutionary WhatsApp-based healthcare companion that:

• Understands voice messages for those who can't type
• Uses advanced AI to offer personalized health insights
• Maintains conversation history for better context
• Suggests home remedies and preventive measures

🌟 Competitive Advantages

Why We're Better Than Existing Solutions

Feature	Meta AI	आरोग्य सेतु (ArogyaSetu)
Voice Messages	❌ No support for voice inputs	✅ Full voice message processing with Indian accent support
Medical Context	❌ Generic responses	✅ Specialized medical knowledge base
Privacy	⚠️ General data policies	✅ HIPAA-compliant, medical-grade security
Response Format	❌ Text-only responses	✅ Structured medical advice with causes, remedies & precautions
Local Context	❌ Global generic responses	✅ India-specific medical guidance
History Tracking	❌ No medical history	✅ Maintains symptom & interaction history
Emergency Handling	❌ No emergency protocols	✅ Emergency detection & escalation
Medication Info	⚠️ Limited information	✅ Common medicine suggestions with dosage
Follow-up Questions	❌ Single response only	✅ Interactive follow-up for better diagnosis
Integration	❌ Standalone chatbot	✅ Future doctor & pharmacy integrations
Cost	💰 Subscription based	🆓 Free for basic healthcare
Availability	⚠️ Region restricted	✅ Available across India

Legend:

• ✅ Full Support
• ⚠️ Partial/Limited
• ❌ Not Available
• 💰 Paid
• 🆓 Free

🚀 Key Features

1. Voice-First Interaction

   • Send voice notes describing symptoms
   • Perfect for elderly users

2. Smart Symptom Analysis

   • Advanced AI-powered diagnosis
   • Contextual follow-up questions
   • Personalized health recommendations

3. Privacy-First Design

   • End-to-end encryption
   • Secure data storage
   • HIPAA-compliant architecture

🛠️ Tech Stack

• Backend: Python, FastAPI
• AI/ML:
  • Whisper ASR for voice recognition
  • Advanced LLMs for medical analysis
  • ChromaDB for conversation memory
• Infrastructure: Docker, SQLite
• API: WhatsApp Cloud API
• Security: JWT, End-to-end encryption

🔄 System Architecture

graph TD
    A[User] -->|WhatsApp Message| B[WhatsApp Cloud API]
    B -->|Webhook| C[FastAPI Backend]
    C -->|Voice| D[Whisper ASR]
    C -->|Text| E[LLM Engine]
    D -->|Transcription| E
    E -->|Query| F[(ChromaDB)]
    F -->|Context| E
    E -->|Response| C
    C -->|Message| B
    B -->|Reply| A

Loading

🧠 AI Pipeline

sequenceDiagram
    participant U as User
    participant W as WhatsApp
    participant B as Backend
    participant ASR as Whisper ASR
    participant AI as LLM Engine
    participant DB as ChromaDB

    U->>W: Send Voice/Text
    W->>B: Forward Message
    alt is voice message
        B->>ASR: Convert to Text
        ASR-->>B: Transcription
    end
    B->>AI: Process Query
    AI->>DB: Get Context
    DB-->>AI: Past Interactions
    AI-->>B: Medical Response
    B->>W: Format Response
    W->>U: Display Result

Loading

📊 Data Flow

flowchart LR
    A[Input] --> B{Type?}
    B -->|Voice| C[ASR Processing]
    B -->|Text| D[Text Processing]
    C --> E[Medical Analysis]
    D --> E
    E --> F[Context Lookup]
    F --> G[Response Generation]
    G --> H[Message Formatting]
    H --> I[User Response]

Loading

📱 How It Works

1. Add ArogyaSetu on WhatsApp: +91-84013-72294
2. Share Your Symptoms:
   • Send a text message OR
   • Record a voice note
3. Get Instant Analysis:
   • Possible causes
   • Recommended remedies
   • Precautions
   • When to see a doctor

🏥 Future Roadmap

1. Q2 2024

   • Medicine reminder integration
   • Doctor appointment booking
   • Lab test recommendations

2. Q3 2024

   • Mental health support
   • Chronic disease management
   • Wellness tracking

3. Q4 2024

   • Telemedicine integration
   • Health insurance guidance
   • Emergency service connection

🌈 Why We'll Win

1. Massive Impact: Serving millions who lack healthcare access
2. Innovation: Voice-first, AI-powered healthcare solution
3. Scalability: WhatsApp's widespread adoption in India
4. Accessibility: No new app downloads needed
5. Privacy: Robust security measures

🏃‍♂️ Quick Start

# Clone the repository
git clone https://github.com/Niharrakholiya/DUhacks-4.0.git

# Install dependencies
pip install -r requirements.txt

# Set up environment variables
cp .env.example .env
# Edit .env with your credentials

# Run the application
uvicorn webhook:app --reload

🔗 Setting Up WhatsApp Webhook

1. Create Meta App:

   • Visit Meta Developers Portal
   • Create new app & select WhatsApp API
   • Get your WhatsApp API Token & Phone Number ID

2. Configure Webhook:

   # Your webhook URL will be:
   https://your-domain.com/webhook/
   
   # Local Testing with Ngrok:
   ngrok http 8000
   # Copy the HTTPS URL provided by ngrok

3. Verify Webhook:

   • Use the ngrok URL in Meta Dashboard
   • Set verify token (same as WHATSAPP_HOOK_TOKEN)
   • Subscribe to messages webhook

4. Test Connection:

   # Send a test message to your WhatsApp number
   # You should receive an automated response

🤝 Team IndustryInnovators - आरोग्य सेतु

• Nihar Rakholiya - AI/ML Lead
• Lakhman Parmar - WhatsApp Integration Lead
• Neil Patel - Text-to-Speech Integration Lead
• Jenish Vamja - Requirements & Deployment Lead

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• WhatsApp for the Cloud API
• DUHacks 4.0 Organizers
• Our mentors and advisors
• Early beta testers

---

Made with ❤️ for a healthier India 🇮🇳