🎤 Sentiment Aura

An AI-powered real-time sentiment analysis application that visualizes emotions through beautiful Perlin noise animations. Speak into your microphone and watch as your words transform into flowing, emotion-driven visual art.

🌐 Live Demo

• Frontend: https://sentiment-aura-eight.vercel.app
• Backend API: https://sentiment-aura-7z75.onrender.com

✨ Features

• 🎙️ Real-time Speech Transcription - Instant audio-to-text using Deepgram API
• 🤖 AI Sentiment Analysis - Powered by Claude Sonnet 4 for emotion detection
• 🌊 Perlin Noise Visualization - Beautiful, organic animations that react to emotions
• 🎨 Emotion-Specific Behaviors - Each emotion has unique colors, speeds, and particle movements
• 📊 Live Sentiment Tracking - Real-time sentiment score (0-100%) and keyword extraction
• 💬 Live Transcript Display - See your words appear in real-time

🛠️ Tech Stack

Frontend

• React - UI framework
• react-p5 - Creative coding for Perlin noise visualization
• Web Audio API - Microphone capture and audio processing
• WebSockets - Real-time communication with Deepgram
• Vercel - Deployment platform

Backend

• FastAPI (Python) - High-performance API framework
• Anthropic Claude API - AI-powered sentiment analysis
• Pydantic - Data validation
• Render - Backend hosting

External APIs

• Deepgram - Real-time speech-to-text transcription
• Claude Sonnet 4 - Natural language understanding and sentiment analysis

🏗️ Architecture

┌─────────────┐         WebSocket          ┌──────────────┐
│   Browser   │ ────────────────────────> │   Deepgram   │
│  (React)    │ <──────────────────────── │     API      │
└─────────────┘      Transcription         └──────────────┘
       │
       │ HTTP POST /process_text
       │ (final transcript)
       ▼
┌─────────────┐                            ┌──────────────┐
│   FastAPI   │ ────────────────────────> │   Claude     │
│   Backend   │ <──────────────────────── │     API      │
└─────────────┘    Sentiment Analysis      └──────────────┘
       │
       │ JSON Response
       │ {sentiment, emotion, keywords}
       ▼
┌─────────────┐
│ Perlin Noise│
│ Visualization│
└─────────────┘

🚀 Local Development Setup

Prerequisites

• Node.js (v16+)
• Python (v3.8+)
• API Keys:
  • Deepgram API Key
  • Anthropic API Key

Backend Setup

# Navigate to backend folder
cd backend

# Create virtual environment
python3 -m venv venv

# Activate virtual environment
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Create .env file
cat > .env << EOF
ANTHROPIC_API_KEY=your_anthropic_key_here
DEEPGRAM_API_KEY=your_deepgram_key_here
EOF

# Start the server
uvicorn main:app --reload --port 8000

Backend will run at: http://localhost:8000

Frontend Setup

# Navigate to frontend folder
cd frontend

# Install dependencies
npm install

# Create .env.local file
cat > .env.local << EOF
REACT_APP_BACKEND_URL=http://localhost:8000
REACT_APP_DEEPGRAM_API_KEY=your_deepgram_key_here
EOF

# Start development server
npm start

Frontend will run at: http://localhost:3000

📦 Deployment

Deploy Backend to Render

1. Push code to GitHub
2. Go to Render Dashboard
3. Create new Web Service
4. Connect your repository
5. Configure:
   • Root Directory: backend
   • Build Command: pip install -r requirements.txt
   • Start Command: uvicorn main:app --host 0.0.0.0 --port $PORT
6. Add environment variables:
   • ANTHROPIC_API_KEY
   • DEEPGRAM_API_KEY

Deploy Frontend to Vercel

1. Go to Vercel Dashboard
2. Import your GitHub repository
3. Configure:
   • Framework Preset: Create React App
   • Root Directory: frontend
   • Build Command: npm run build
4. Add environment variables:
   • REACT_APP_BACKEND_URL: Your Render backend URL
   • REACT_APP_DEEPGRAM_API_KEY: Your Deepgram key
5. Deploy!

🎮 Usage

1. Open the app in your browser
2. Click "Start Recording" button
3. Allow microphone access when prompted
4. Speak clearly into your microphone
5. Watch the magic happen:
   • Your words appear as live transcript
   • Background changes color based on emotion
   • Particle speed reflects sentiment intensity
   • Keywords are extracted and displayed

Emotion Examples

• 😊 Happy: "I'm so happy and grateful today!"

  • Bright yellow/gold particles, medium-fast movement

• 🤩 Excited: "This is amazing! I'm so excited!"

  • Hot pink/magenta, fast chaotic sparkles

• 😢 Sad: "I feel sad and lonely"

  • Deep blue, slow drifting particles

• 😠 Angry: "I'm really frustrated and angry!"

  • Pulsing red background, aggressive movement

• 😌 Calm: "I feel peaceful and calm"

  • Smooth cyan/aqua, gentle floating

📁 Project Structure

sentiment-aura/
├── backend/
│   ├── main.py              # FastAPI application
│   ├── requirements.txt     # Python dependencies
│   └── .env                 # Environment variables (not in repo)
│
├── frontend/
│   ├── public/
│   ├── src/
│   │   ├── App.js           # Main component
│   │   ├── AudioCapture.js  # Microphone & Deepgram integration
│   │   └── AuraVisualization.js  # p5.js Perlin noise visualization
│   ├── package.json
│   └── .env.production      # Production environment variables
│
├── .gitignore
└── README.md

🔌 API Endpoints

GET /

Health check endpoint

Response:

{
  "message": "Sentiment Aura Backend Running!"
}

POST /process_text

Analyze sentiment and extract keywords from text

Request:

{
  "text": "I am so happy and excited!"
}

Response:

{
  "sentiment": 0.9,
  "emotion": "excited",
  "keywords": ["happy", "excited"]
}

🎨 Visualization Details

The Perlin noise visualization uses organic, flowing particle systems that respond to emotional data:

• Color Hue: Determined by emotion type
• Saturation & Brightness: Mapped to sentiment score (0-1)
• Particle Speed: Higher sentiment = faster movement
• Noise Scale: Controls smoothness (calm = very smooth, angry = chaotic)
• Connections: Lines between nearby particles create flowing patterns

Emotion-Specific Parameters

Emotion	Color	Speed	Behavior
Happy	Yellow/Gold	Medium-Fast	Sparkles
Excited	Pink/Magenta	Very Fast	Chaotic sparkles
Sad	Deep Blue	Slow	Heavy drift
Angry	Red	Fast	Pulsing, aggressive
Calm	Cyan/Aqua	Very Slow	Smooth floating
Neutral	Muted Green	Medium	Balanced flow

🐛 Troubleshooting

Microphone not working

• Ensure you're using HTTPS or localhost
• Check browser permissions for microphone access
• Try a different browser (Chrome recommended)

CORS errors

• Verify backend CORS settings include your frontend URL
• Check that backend is running and accessible

Sentiment not updating

• Check browser console for errors
• Verify backend environment variables are set
• Test backend endpoint directly with curl

Backend spinning down (Render free tier)

• First request after inactivity takes 30-60 seconds
• Subsequent requests are fast

📝 License

MIT License - feel free to use this project for learning or personal projects!

🙏 Acknowledgments

• Deepgram - Real-time speech transcription
• Anthropic - Claude AI for sentiment analysis
• p5.js - Creative coding framework
• Memory Machines - Project inspiration

👨‍💻 Author

Archit Gupta

• GitHub: @archit2901
• Project: sentiment-aura

---

⭐ Star this repo if you found it helpful!