WhatsApp AI Math Assistant

This project integrates a WhatsApp Business API with an AI-powered backend to handle math-related queries in both Arabic and English. It features:

• Language Detection & Translation: Automatically detects the message language and translates Arabic to English using the OpenL Free API.
• Math Evaluation: Utilizes Sympy for evaluating mathematical expressions.
• AI Agent: Employs OpenAI's GPT-3.5-Turbo via LangGraph to determine if a query is math-related.
• Voice Response: Sends a predefined Arabic audio response for voice messages using ElevenLabs TTS.
• Webhook Integration: FastAPI server with GET and POST endpoints to handle WhatsApp webhook verification and message processing.

---

📂 Repository Structure

AlsunAI-Task/
├── math_agent.py # LangGraph agent setup with GPT-3.5-Turbo
├── translation.py # Language detection and translation logic
├── webhook.py # FastAPI server with webhook endpoints
├── tools/
│ └── calculator.py # Sympy-based calculator tool
├── static/
│ └── Arabic.mp3 # Predefined Arabic audio response
├── requirements.txt # Python dependencies
├── README.md # Project documentation
└── Demo.mp4 # Demonstration video

---

🚀 Getting Started

Prerequisites

• Python 3.8 or higher
• Ngrok (for tunneling localhost)
• WhatsApp Business API credentials
• OpenAI API key
• OpenL Free API credentials

Installation

1. Clone the repository:

   git clone https://github.com/MrSa3dola/AlsunAI-Task.git
   cd AlsunAI-Task```

2. Install dependencies:

pip install -r requirements.txt 3. Set up environment variables:

Create a .env file in the root directory and add your API keys:

OPENAI_API_KEY=
OPENL_API_KEY=
VERSION=
WA_TOKEN=
PHONE_NUMBER_ID=
WA_ACCOUNT_ID=
WA_VERIFY_TOKEN=

4. Run the FastAPI server: uvicorn webhook:app --reload
5. Expose the server using Ngrok: ngrok http 8000 Use the generated Ngrok URL to set up your WhatsApp webhook.

Demo

Link

Documentation

Link