WhatsApp AI Backend

A comprehensive WhatsApp Business API backend with AI-powered image analysis capabilities using OpenAI and Anthropic.

🚀 Features

• WhatsApp Business API Integration: Send/receive messages via WhatsApp webhook
• AI-Powered Image Analysis: Analyze images using OpenAI GPT-4 Vision and Anthropic Claude-3
• Multi-Provider AI Support: Switch between OpenAI and Anthropic for different use cases
• File Upload & Processing: Upload and process images with automatic resizing
• Rate Limiting: Built-in rate limiting for API protection
• Comprehensive Logging: Winston-based logging with different levels
• Health Monitoring: Health check endpoints for monitoring
• Type-Safe: Full TypeScript implementation with strict typing

📋 Prerequisites

• Node.js 18+
• TypeScript
• WhatsApp Business Account with API access
• OpenAI API Key
• Anthropic API Key

🛠️ Installation

1. Clone and navigate to the project:

cd whatsapp-ai-backend

2. Install dependencies:

npm install

3. Copy environment configuration:

cp .env.example .env

4. Configure your environment variables in .env:

# Server Configuration
PORT=3000
NODE_ENV=development

# AI Services
OPENAI_API_KEY=your_openai_api_key_here
ANTHROPIC_API_KEY=your_anthropic_api_key_here

# WhatsApp Business API
WHATSAPP_API_URL=https://graph.facebook.com/v18.0
WHATSAPP_ACCESS_TOKEN=your_whatsapp_access_token_here
WHATSAPP_PHONE_NUMBER_ID=your_phone_number_id_here
WHATSAPP_VERIFY_TOKEN=your_verify_token_here

# File Upload
MAX_FILE_SIZE=10485760
UPLOAD_PATH=./uploads

# Logging
LOG_LEVEL=info
LOG_FILE=app.log

# Rate Limiting
RATE_LIMIT_WINDOW_MS=900000
RATE_LIMIT_MAX=100

🚀 Usage

Development

npm run dev

Production

npm run build
npm start

Testing

npm test

Linting & Type Checking

npm run lint
npm run typecheck

📡 API Endpoints

WhatsApp Endpoints

GET /api/whatsapp/webhook

Webhook verification for WhatsApp Business API.

POST /api/whatsapp/webhook

Webhook endpoint for receiving WhatsApp messages.

POST /api/whatsapp/send-message

Send messages via WhatsApp Business API.

Body:

{
  "to": "+5511999999999",
  "message": "Hello from AI!",
  "type": "text"
}

AI Endpoints

POST /api/ai/analyze

Analyze images or text using AI.

Body:

{
  "imageUrl": "https://example.com/image.jpg",
  "text": "Optional text to analyze",
  "provider": "openai",
  "options": {
    "prompt": "Analyze this image",
    "temperature": 0.7,
    "max_tokens": 500
  }
}

POST /api/ai/analyze-upload

Upload and analyze an image.

Form Data:

• image: Image file
• prompt: Analysis prompt
• provider: AI provider (openai/anthropic)

POST /api/ai/generate

Generate text responses.

Body:

{
  "prompt": "Write a haiku about technology",
  "provider": "anthropic"
}

GET /api/ai/providers

Get available AI providers and their capabilities.

Health Endpoints

GET /api/health

Get application health status.

GET /api/health/ready

Check if application is ready to serve requests.

🤖 WhatsApp Bot Features

The bot automatically responds to different types of messages:

• Text Messages: Generates AI-powered responses
• Images: Analyzes images and provides detailed descriptions
• Greetings: Responds with welcome message in Portuguese/English
• Help Requests: Provides usage instructions

Supported Commands

• Send "hello", "hi", or "oi" for greetings
• Send "help" or "ajuda" for assistance
• Send images for AI analysis
• Use keywords like "analyze" or "describe" for detailed analysis

🏗️ Project Structure

src/
├── types/           # TypeScript type definitions
├── services/        # Business logic services
│   ├── aiService.ts     # AI integration (OpenAI + Anthropic)
│   └── whatsappService.ts # WhatsApp Business API
├── routes/          # API route handlers
│   ├── ai.ts           # AI-related endpoints
│   ├── whatsapp.ts     # WhatsApp webhook endpoints
│   └── health.ts       # Health check endpoints
├── middleware/      # Express middleware
│   ├── errorHandler.ts # Error handling
│   └── rateLimiter.ts  # Rate limiting
├── utils/           # Utility functions
│   ├── config.ts       # Environment configuration
│   └── logger.ts       # Winston logging setup
└── index.ts         # Application entry point

🔧 Configuration

AI Providers

• OpenAI: GPT-4 Vision for image analysis, GPT-3.5 Turbo for text
• Anthropic: Claude-3 Sonnet for image analysis, Claude-3 Haiku for text

File Upload

• Maximum file size: 10MB (configurable)
• Supported formats: JPEG, PNG, GIF, WebP
• Automatic image resizing to 1024x1024 max
• Files stored in ./uploads directory

Rate Limiting

• Default: 100 requests per 15 minutes per IP
• Configurable via environment variables

📊 Monitoring

Logging

• Winston-based logging with multiple transports
• Separate error log file
• Structured JSON logging for production
• Console logging for development

Health Checks

• /api/health: Application status
• /api/health/ready: Readiness probe
• Memory usage monitoring
• Uptime tracking

🔒 Security Features

• Helmet: Security headers
• CORS: Cross-origin resource sharing
• Rate Limiting: Request throttling
• Input Validation: Joi schema validation
• Error Handling: Secure error responses
• File Upload Security: Type and size validation

🐳 Docker Support (Optional)

Create Dockerfile:

FROM node:18-alpine
WORKDIR /app
COPY package*.json ./
RUN npm ci --only=production
COPY dist ./dist
EXPOSE 3000
CMD ["npm", "start"]

📝 License

MIT License - See LICENSE file for details.

🤝 Contributing

1. Fork the repository
2. Create your 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

📞 Support

For support, create an issue in the GitHub repository.