Abstract

This project implements an intelligent chatbot powered by Microsoft's DialoGPT, a state-of-the-art conversational AI model, to enable natural and human-like interactions. The chatbot is integrated with WhatsAuto, a WhatsApp automation tool, allowing it to send and receive messages automatically on the WhatsApp platform. Built using the Flask framework, the system provides a seamless backend for handling user queries and generating context-aware responses. This solution is designed for applications such as customer support, FAQs, and personalized messaging, offering a scalable and efficient way to automate conversations on one of the world's most popular messaging platforms.

Methodology

The implementation of the chatbot involves the following key steps and technologies:

1. Chatbot Development with DialoGPT

• DialoGPT Model: Microsoft's DialoGPT, a pre-trained conversational AI model based on GPT architecture, is used to generate human-like responses.
• Fine-Tuning (Optional): The model can be fine-tuned on custom datasets to improve its performance for specific use cases.
• Response Generation: User inputs are processed by DialoGPT, which generates contextually relevant responses.

2. Backend Development with Flask

• Flask Framework: A lightweight Python web framework is used to create the backend server for the chatbot.
• API Endpoints: RESTful APIs are implemented to handle incoming requests, process user inputs, and return chatbot responses.
• Integration with DialoGPT: The Flask backend communicates with the DialoGPT model to generate responses dynamically.

3. WhatsApp Automation with WhatsAuto

• WhatsAuto Integration: The chatbot is integrated with WhatsAuto, a tool for automating WhatsApp messaging.
• Message Handling: WhatsAuto listens for incoming messages on WhatsApp, forwards them to the Flask backend, and sends the chatbot's responses back to the user.
• Automation Workflow: The system ensures seamless two-way communication between the user and the chatbot on WhatsApp.

4. System Workflow

1. A user sends a message on WhatsApp.
2. WhatsAuto captures the message and forwards it to the Flask backend through Ngrok.
3. The Flask backend processes the message using DialoGPT to generate a response.
4. The response is sent back to WhatsAuto, which delivers it to the user on WhatsApp.

View bot.py on GitHub

Results