ConversAI MVP: AI Driven Digital Human

A voice-first conversational AI application that enables real-time voice interaction with an AI agent using Microsoft's DialoGPT model.

🎯 Project Overview

This Semester 7 MVP validates the core concept of voice-driven AI conversation with the following key features:

• Voice Input: Web Speech API for real-time speech-to-text
• AI Processing: Microsoft DialoGPT-small for conversation generation
• Voice Output: Browser Speech Synthesis for text-to-speech
• Conversation History: SQLite database for persistent chat storage
• Modern UI: Responsive web interface with real-time feedback

🚀 Quick Start

Prerequisites

• Python 3.8+ (recommended: Python 3.10+)
• Modern web browser with Web Speech API support (Chrome, Edge, Safari)
• Microphone access for voice input

Installation & Setup

1. Clone and navigate to the project:

   git clone <repository-url>
   cd ConversAI-MVP

2. Create and activate virtual environment:

   python -m venv venv
   
   # On Windows:
   venv\Scripts\activate
   
   # On macOS/Linux:
   source venv/bin/activate

3. Install backend dependencies:

   cd backend
   pip install -r requirements.txt

4. Initialize the database:

   python database.py

5. Test the model (optional but recommended):

   python test_model.py

6. Start the backend server:

   python app.py

7. Open the frontend:

   • Navigate to http://localhost:5000 in your browser
   • Or open frontend/index.html directly in your browser

🎮 Usage

Voice Mode (Recommended)

1. Click the microphone button 🎤
2. Speak clearly into your microphone
3. Wait for the AI to process and respond
4. The AI will respond both in text and voice

Text Mode (Fallback)

1. If voice is not supported, use the text input box
2. Type your message and press Enter or click Send
3. Receive text and voice responses

🏗️ Architecture

Frontend (HTML/CSS/JS)
├── Web Speech API (STT)
├── Chat Interface
└── Speech Synthesis (TTS)

Backend (Python/Flask)
├── REST API (/api/chat)
├── DialoGPT Model
└── SQLite Database

Data Flow:
Browser Voice → STT → Backend → DialoGPT → Response → TTS → Browser Voice

📁 Project Structure

ConversAI-MVP/
├── frontend/
│   ├── index.html          # Main UI
│   ├── style.css           # Styling
│   └── script.js           # Web Speech API & chat logic
├── backend/
│   ├── app.py              # Flask server
│   ├── model.py            # DialoGPT integration
│   ├── database.py         # SQLite operations
│   ├── test_model.py       # Model testing
│   └── requirements.txt    # Python dependencies
├── .gitignore
└── README.md

🔧 API Endpoints

• GET / - Serves the frontend
• POST /api/chat - Main chat endpoint
  • Input: {"message": "user input", "session_id": "optional"}
  • Output: {"reply": "bot response", "session_id": "session_id"}
• GET /api/history/<session_id> - Get conversation history
• GET /api/health - Health check

🧪 Testing

Model Testing

cd backend
python test_model.py

Manual Testing Checklist

• Backend starts without errors
• Frontend loads in browser
• Microphone button responds to clicks
• Voice input is captured and transcribed
• AI generates relevant responses
• Text-to-speech works
• Conversation history is saved to database
• Multiple conversation turns work
• Error handling works (try speaking when mic is off)

⚙️ Configuration

Model Parameters

Edit backend/model.py to adjust:

• max_length: Maximum response length
• temperature: Response creativity (0.1-1.0)
• top_k: Vocabulary diversity
• top_p: Nucleus sampling

Speech Recognition

Edit frontend/script.js to adjust:

• recognition.lang: Language setting
• utterance.rate: Speech speed
• utterance.pitch: Voice pitch

🐛 Troubleshooting

Common Issues

"Model not loading"

• Ensure you have sufficient RAM (2GB+ recommended)
• Check internet connection for model download
• Try running python test_model.py to diagnose

"Speech recognition not working"

• Check microphone permissions in browser
• Use HTTPS or localhost (required for Web Speech API)
• Try Chrome or Edge browser

"Backend connection failed"

• Ensure Flask server is running on port 5000
• Check for firewall blocking
• Verify CORS settings in app.py

"Empty responses"

• Check model loading in backend logs
• Try simpler inputs first
• Verify database is initialized

Performance Tips

• First run: Model download may take 5-10 minutes
• Response time: 2-6 seconds for local inference
• Memory usage: ~2GB RAM for DialoGPT-small
• Browser: Chrome/Edge recommended for best Web Speech API support

📊 Success Metrics

• ✅ Voice input captured and transcribed accurately
• ✅ AI responses generated within 2-6 seconds
• ✅ Text-to-speech working smoothly
• ✅ Conversation history persisted
• ✅ Multiple conversation turns maintained
• ✅ Error handling graceful

🔮 Future Enhancements

Semester 8+ Roadmap

• Avatar Integration: 3D/2D digital human rendering
• Advanced Models: GPT-3.5/4, Claude, or local fine-tuned models
• Multi-language Support: Internationalization
• Emotion Detection: Voice emotion analysis
• Vector Database: Pinecone for long-term memory
• Real-time Streaming: WebSocket for faster responses
• Mobile App: React Native or Flutter
• Cloud Deployment: AWS/Azure with auto-scaling

Technical Upgrades

• Backend: FastAPI for better performance
• Frontend: React/Vue.js for complex UI
• Database: PostgreSQL for production scale
• Caching: Redis for session management
• Monitoring: Prometheus + Grafana
• CI/CD: GitHub Actions for automated testing

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📝 License

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

👥 Team

• Product Manager: ConversAI MVP Team
• Technical Lead: ConversAI MVP Team
• Code Generator: Cursor AI Assistant

📞 Support

For issues and questions:

1. Check the troubleshooting section above
2. Review the test results from python test_model.py
3. Check browser console for frontend errors
4. Check backend logs for server errors

---

Happy Conversing! 🤖💬

Built with ❤️ for Semester 7 MVP validation