Jinx: Chop Now Restaurant Chatbot

Overview

This project implements a chatbot(Jinx) for Chop Now, an hypothetical restaurant located in Lagos, Nigeria. The chatbot is designed to assist users in placing new food orders, tracking existing orders, and answering queries. Built using Dialogflow for natural language processing, FastAPI for backend services, and MySQL for data storage, this project demonstrates an end-to-end solution for an AI-driven customer service system.

The bot supports both English and Pidgin, catering to local Nigerian users, making it more accessible and engaging.

Demo

Watch the demo of the chatbot in action here.

Problem Statement

Jinx aims to enhance its customer service by providing an easy-to-use chatbot that allows users to:

• Place food orders through natural conversations.
• Track their order statuses without needing to call customer support.
• Get immediate responses to FAQs, such as menu options and restaurant location.

Architecture

The system architecture can be broken down into several components:

1. Dialogflow:

   • Handles natural language understanding (NLU) and processes user intents such as placing or tracking orders.
   • Manages conversational context, enabling the bot to provide relevant responses based on user history.

2. FastAPI Backend:

   • Acts as a middleman between Dialogflow and the MySQL database.
   • Receives user inputs (order details, queries) from Dialogflow and processes them, interacting with the database to store or retrieve information.

3. MySQL Database:

   • Stores customer orders, menu items, and other relevant data.
   • The backend fetches data for order tracking or new orders and updates the database accordingly.

4. Ngrok:

   • Facilitates secure HTTPS tunneling to expose the local FastAPI server to the web for testing with Dialogflow.

Project Structure

├── backend/                  # Contains FastAPI backend code
│   ├── main.py               # FastAPI entry point
│   └── requirements.txt      # Backend dependencies
├── db/                       # Contains MySQL database dump
│   └── chopnow_db.sql        # Database dump file
├── dialogflow_assets/         # Dialogflow training phrases, intents, etc.
└── README.md                 # Project documentation
└── food list.png              #png file was uploaded to a google drive and then linked in dialogflow

Features

• New Order: Users can place food orders by selecting from a predefined menu.
• Order Tracking: Users can track their order status using an order ID.
• Language Support: The chatbot supports both English and Nigerian Pidgin, providing localized conversations for users.
• Menu Querying: The chatbot allows users to inquire about the menu and dish availability.

Installation

Prerequisites

• Python
• MySQL server
• Dialogflow (Google Cloud)
• Ngrok (for local testing)
• mysql-connector-python
• fastapi[all]

Usage

• Open Dialogflow and start a conversation with the chatbot to test intents such as placing an order or tracking an existing one.
• You can also interact with the FastAPI backend directly by making requests to the API endpoints exposed by the server.

Key Learnings

This project demonstrates the integration of multiple technologies to build a complete chatbot system:

• Dialogflow for natural language understanding and intent recognition.
• FastAPI for a lightweight, fast backend service.
• MySQL for efficient data storage and retrieval.
• Ngrok for local development.

Future Enhancements

• Payment Integration: Add support for online payments via the chatbot.
• Multiple Language Support: Extend support for more Nigerian languages.
• Extended Order Features: Enable users to modify or cancel their orders.