README.md

metadata

title: PropAgent
emoji: 🤖
colorFrom: blue
colorTo: purple
sdk: docker
pinned: false
short_description: WhatsApp AI Agent with Supabase integration

PropAgent - WhatsApp AI Agent

A WhatsApp chatbot powered by OpenAI GPT-4 and integrated with Supabase for user management.

Features

• 🤖 AI-powered chat responses using OpenAI GPT-4
• 📱 WhatsApp Business API integration
• 👥 User management with Supabase database
• 🔄 Conversation memory and context
• 📊 User activity tracking
• 🚀 RESTful API endpoints for user management

Setup

1. Environment Variables

Create a .env file with the following variables:

# WhatsApp Business API
VERIFY_TOKEN=your_webhook_verify_token
PHONE_NUMBER_ID=your_phone_number_id
WHATSAPP_API_TOKEN=your_whatsapp_api_token

# OpenAI
OPENAI_API_KEY=your_openai_api_key

# Supabase
SUPABASE_URL=your_supabase_project_url
SUPABASE_KEY=your_supabase_anon_key

2. Supabase Setup

1. Create a new Supabase project at supabase.com
2. Go to the SQL Editor in your Supabase dashboard
3. Run the SQL script from supabase_setup.sql to create the users table
4. Copy your project URL and anon key from Settings > API

3. WhatsApp Business API Setup

1. Create a Meta Developer account
2. Set up a WhatsApp Business app
3. Configure webhook URL: https://your-domain.com/webhook
4. Set the verify token in your environment variables

4. Installation

pip install -r requirements.txt

5. Running the Application

uvicorn main:app --host 0.0.0.0 --port 8000

API Endpoints

Webhook Endpoints

• GET /webhook - WhatsApp webhook verification
• POST /webhook - Receive WhatsApp messages

User Management

• GET /users/{wa_id} - Get user by WhatsApp ID
• GET /users - List all users (with pagination)
• PUT /users/{wa_id} - Update user name

Health & Testing

• GET /ping - Simple health check
• GET /health - Detailed service status
• POST /chat - Direct AI chat endpoint

Database Schema

Users Table

users
-----
- wa_id (TEXT, PRIMARY KEY) - WhatsApp user ID
- name (TEXT) - User's display name
- created_at (TIMESTAMP) - Account creation time
- updated_at (TIMESTAMP) - Last activity time

Features

User Management

• Automatic user creation when first message is received
• User name updates from WhatsApp profile
• Activity tracking with automatic timestamp updates

AI Chat

• Context-aware conversations using user names
• Conversation memory (last 6 messages)
• Error handling and fallback responses

Security

• Environment variable configuration
• Optional Row Level Security (RLS) in Supabase
• Input validation and error handling

Project Structure

PropAgent/
├── main.py              # Main application entry point
├── config.py            # Configuration and client setup
├── database.py          # Supabase database operations
├── whatsapp.py          # WhatsApp API integration
├── ai_chat.py           # LangGraph AI conversation logic
├── api_routes.py        # FastAPI route definitions
├── supabase_setup.sql   # Database schema setup
└── requirements.txt     # Python dependencies

Development

The application uses a modular architecture:

• main.py - Clean entry point focusing on core webhook logic
• config.py - Environment variables and client initialization
• database.py - All Supabase database operations
• whatsapp.py - WhatsApp Business API integration
• ai_chat.py - LangGraph conversation flow and AI processing
• api_routes.py - REST API endpoints and webhook verification

Technologies:

• FastAPI for the web framework
• LangGraph for conversation flow management
• OpenAI GPT-4 for AI responses
• Supabase for user data storage
• WhatsApp Business API for messaging

Deployment

The application includes a Dockerfile for containerized deployment. You can deploy it to:

• Hugging Face Spaces
• Railway
• Heroku
• Any container platform

License

MIT License