Jinx: AI-Powered Chatbot for Chop Now Restaurant

Abstract

Jinx is an AI-powered chatbot developed for Chop Now, a hypothetical restaurant in Lagos, Nigeria. It enhances customer service by allowing users to:

• Place food orders naturally via conversation.
• Track order statuses without calling support.
• Get instant responses to FAQs about the menu and restaurant location.

Built with Dialogflow (NLP), FastAPI (backend services), and MySQL (data storage), Jinx provides a seamless, multilingual (English & Nigerian Pidgin) interaction experience. This publication details the system’s architecture, implementation, and future improvements.

---

1. Introduction

1.1 Problem Statement

Restaurants often struggle with managing high volumes of customer queries and orders. Traditional phone-based support is inefficient, leading to:

• Long wait times for order placement or tracking.
• Miscommunications due to language barriers.
• Overloaded customer service agents, reducing efficiency.

Jinx solves these issues by automating food ordering, tracking, and inquiries using conversational AI.

1.2 Objectives

Jinx is designed to:

• Reduce customer support dependency by handling order-related queries automatically.
• Support natural and intuitive ordering in English and Nigerian Pidgin.
• Provide instant responses for menu inquiries and order tracking.

1.3 Current State & Gap Identification

Traditional food ordering relies heavily on:

• Call centers that are prone to delays and human errors.
• Mobile apps that require downloads and manual input.
• Walk-in orders, which limit convenience for customers.

Jinx bridges this gap by offering an instant, AI-driven chatbot that simplifies the ordering process.

---

2. System Architecture

2.1 Overview

The chatbot is built using a modular architecture for scalability and efficiency:

• Dialogflow: Handles intent recognition and natural language processing (NLP).
• FastAPI: Manages API requests, processes orders, and interacts with the database.
• MySQL: Stores orders, menu data, and customer interactions.
• Ngrok: Provides a secure HTTPS tunnel for testing with Dialogflow.

2.2 Architecture Diagram

2.3 Key Components

(A) Dialogflow (Natural Language Processing)

• Understands user queries and extracts intents.
• Maintains conversational context to handle ongoing orders.

(B) FastAPI Backend

• Acts as middleware between Dialogflow and MySQL.
• Processes order placements and tracking requests.

(C) MySQL Database

• Stores customer orders, menu items, and tracking data.
• Allows efficient retrieval and updates for order tracking.

(D) Ngrok (For Local Testing)

• Exposes the local FastAPI server via a public HTTPS URL.

---

3. Implementation

3.1 Project Structure

├── backend/                  # FastAPI backend code
│   ├── main.py               # API entry point
│   ├── db_helper.py          # Handles database operations
│   ├── generic_helper.py     # Utility functions
│   ├── requirements.txt      # Dependencies
├── db/                       # MySQL database dump
│   ├── chop_now_restaurant.sql # Database schema
├── dialogflow_assets/        # Intents and training data
├── static/                   # Static assets (e.g., menu images)
└── README.md                 # Documentation

3.2 Dataset Sources & Collection

The chatbot utilizes a structured dataset containing:

• Menu items & pricing from the Chop Now restaurant.
• Customer orders & transaction history for personalized recommendations.
• Order tracking statuses to provide real-time updates.

Data is collected through manual entry and API-based updates from the restaurant's inventory system.

3.3 Dataset Processing Methodology

• Normalization: Cleaning and structuring data for consistency.
• Indexing: Optimizing queries for fast retrieval.
• Encryption: Secure handling of customer order details.

---

4. Installation & Setup

4.1 Prerequisites

• Python 3.x
• MySQL Server
• Google Dialogflow Account
• Ngrok (for local testing)

4.2 Installation Steps

Step 1: Clone the Repository

git clone https://github.com/monsterdevgit/Chop_Now_Restaurant_Chatbot
cd chopnow-chatbot

Step 2: Setup MySQL Database

CREATE DATABASE chop_now_restaurant;
USE chop_now_restaurant;
SOURCE db/chop_now_restaurant.sql;

Step 3: Run FastAPI Backend

uvicorn backend.main:app --reload

Step 4: Connect FastAPI to Dialogflow

• Use Ngrok to expose the local FastAPI server:

ngrok http 8000

• Copy the HTTPS URL and update Dialogflow's fulfillment settings.

---

5. Performance & Maintenance

5.1 Performance Metrics Analysis

5.2 Maintenance Considerations

• Database backups every 24 hours.
• Periodic NLP training to enhance accuracy.
• Server health monitoring via logging & alerts.

---

6. Access & Licensing

• Source Code Repository: GitHub
• License: Open-source (MIT License)
• Technical Documentation: API Docs

---

7. Contact & Support

• Maintainer: Kenneth Essien
• Email: kennethessien77@gmail.com
• Support Status: Not Actively maintained, community contributions welcome.

---

8. Industry Insights & Future Enhancements

• AI-driven chatbots are reducing customer wait times by 40%.
• Restaurants using AI see increased efficiency & reduced operational costs.
• Future upgrades include payment integration & additional language support.

---

9. Conclusion

Jinx successfully demonstrates the power of AI-driven automation in the restaurant industry. By integrating NLP, backend processing, and database management, this project improves customer service efficiency while reducing human dependency. Future improvements will expand payment integration and language support.