Whatsapp-Bot

This WhatsApp bot automates tasks using the Meta API and OpenAI. It responds intelligently, sets reminders, fetches data, and interacts with APIs. Features include AI-powered replies, task automation, and custom commands. Ideal for personal assistance, business automation, and education, it enhances productivity with smooth, secure communication.

Required Tools and Services

• Twilio Account: Required for accessing the WhatsApp API.
• Flask Server: Used to handle incoming messages and respond via Twilio.
• Ngrok or Similar: Needed to expose your local Flask server to the internet for Twilio to communicate.

Setup Instructions

1. Create a .env file with the following content:

OPENAI_API_KEY=your_openai_api_key
TWILIO_ACCOUNT_SID=your_twilio_account_sid
TWILIO_AUTH_TOKEN=your_twilio_auth_token

2. Install required packages:

pip install openai python-dotenv twilio flask

3. Configure Twilio WhatsApp Sandbox:

• Go to the Twilio Console
• Navigate to Messaging > Try it > WhatsApp
• Set up the WhatsApp sandbox
• Configure the webhook URL to point to your server: https://your-domain.com/bot

4. Run the server locally and expose it using ngrok:

# In one terminal
python whatsapp_bot.py

# In another terminal
ngrok http 5000

• Update the Twilio webhook URL with the ngrok URL.

Configuration File (saved_config.json)

The bot uses a configuration file named saved_config.json, which stores essential settings. Ensure this file is present in your project directory. Below is the default structure:

{
  "openai_model": "gpt-3.5-turbo",  
  "max_history_length": 10,  
  "default_response": "I'm here to help! Ask me anything.",
  "allowed_commands": ["!clear", "!help", "!info"],
  "admin_users": []  
}

• openai_model: Defines which GPT model the bot will use.
• max_history_length: Limits stored conversation history per user.
• default_response: The message the bot sends if no meaningful response is generated.
• allowed_commands: Specifies available commands.
• admin_users: Stores admin user details (currently empty by default).

How to Use the Bot

Once set up, users can interact with the WhatsApp bot by sending messages. The bot supports:

• Regular AI-powered responses
• Special Commands:
  • !clear - Clear conversation history
  • !help - Show help menu
  • !info - Show bot configuration

The bot keeps track of conversation history for each user, allowing for contextual responses.

Notes

• Ensure your Twilio account is properly set up to send and receive WhatsApp messages.
• Use ngrok or deploy the Flask app to a cloud server to make it accessible.
• Keep your API keys and tokens secure by using environment variables.