Build AI WhatsApp Bots with Pure Python

In this guide, we'll explore how to build a WhatsApp bot using Meta’s Cloud API with pure Python and Flask. We’ll also dive into integrating webhook events to receive messages in real-time and using OpenAI for AI-driven responses.

---

Prerequisites

• A Meta developer account. Create one here.

• A business app. Learn to create one here.

• Familiarity with Python.

Table of Contents

1. Get Started

2. Step 1: Select Phone Numbers

3. Step 2: Send Messages with the API

4. Step 3: Configure Webhooks to Receive Messages

5. Step 4: Understanding Webhook Security

6. Step 5: Learn about the API and Build Your App

7. Step 6: Integrate AI into the Application

8. Step 7: Add a Phone Number

Get Started

In this section, we’ll begin setting up your app environment, integrating the WhatsApp API, and preparing for real-time communication via webhook events.

Step 1: Select Phone Numbers

To get started, you'll need to add a phone number to your Meta app.

• Go to the API Setup in your Meta Developer dashboard.

• Choose a test number for sending messages to up to 5 WhatsApp contacts, including your own.

• Verify your phone number by entering the code sent to your WhatsApp app.

Step 2: Send Messages with the API

With a test number ready, you can send your first message via the WhatsApp API.

• Obtain a 24-hour access token from the API access section.

• Convert the provided curl example into a Python function using the requests library.

• Update your .env file with necessary environment variables like APP_ID, APP_SECRET, ACCESS_TOKEN, etc.

• Send a "Hello World" message. You should see the message within 60–120 seconds.

For more persistent access, set up a System User to generate longer-lived access tokens (e.g., 60 days or non-expiring).

Step 3: Configure Webhooks to Receive Messages

Webhook configuration is crucial for receiving messages in real time.

• Install required dependencies (pip install -r requirements.txt) and start your Flask app.

• Use ngrok to expose your local environment to the web for webhook testing.

  • Download the ngrok agent and authenticate with your token.

  • Start ngrok with ngrok http 8000 --domain your-domain.ngrok-free.app.

• In the Meta App Dashboard, go to WhatsApp > Configuration.

  • Enter your ngrok URL followed by /webhook in the Callback URL field.

  • Set a verification token of your choice and ensure it matches the environment variable VERIFY_TOKEN.

Test your setup by sending a message to your app, ensuring it logs and responds appropriately.

Step 4: Understanding Webhook Security

Meta ensures that webhook requests are secure through verification processes.

• Verification Requests: When a webhook is first configured, Meta sends a GET request that must be verified by responding with the hub.challenge value.

• Validating Payloads: All event notifications are signed using SHA256. Your app should validate this signature to ensure data integrity.

---

Step 5: Learn about the API and Build Your App

Familiarize yourself with the WhatsApp API by referring to the official documentation. Here, you’ll learn how to send various types of messages, including text, images, and template messages.

---

Step 6: Integrate AI into the Application

Now that your WhatsApp bot is functioning, we can make it more intelligent by integrating AI.

• Create a function generate_response() using OpenAI’s API.

• Update the environment with OPENAI_API_KEY and OPENAI_ASSISTANT_ID.

• Modify whatsapp_utils.py to integrate your AI response mechanism, so the bot can generate dynamic replies based on user inputs.

---

Step 7: Add a Phone Number

When you’re ready for production, add a dedicated phone number to send messages.

• You can either migrate an existing WhatsApp number to the business platform or use a new number.

• Options for testing include using virtual phone numbers, dual SIMs, or temporary number services.

Ensure that your chosen number is correctly configured in the WhatsApp Business API to avoid service disruptions.

---

By following this guide, you’ll be able to set up a fully functional AI-powered WhatsApp bot using Python, Flask, and the Meta Cloud API. Now you can build, test, and deploy creative chatbot applications for various use cases!