🤖 ChatBot Project — Flask + React + Machine Learning

A full-stack chatbot application built using Flask, React (Vite), and a machine-learning–based intent classification model.
The chatbot uses Natural Language Processing (NLP) with NLTK and a neural network trained using TensorFlow/Keras to generate intent-based responses.

The frontend provides a clean, ChatGPT-style user interface, while the backend handles both the chatbot API and serving the production frontend.

---

🚀 Live Demo

🔗 Deployed on Render
https://chatbot-001-xyz.onrender.com/

⚠️ Note: On Render’s free tier, the first request after inactivity may take a few seconds due to cold start.

---

✨ Features

• Intent-based chatbot using Machine Learning
• NLP preprocessing with NLTK (tokenization + stemming)
• Neural network trained using TensorFlow/Keras
• ChatGPT-style UI with dark/light theme toggle
• FAQ quick-prompt buttons
• Markdown support for bot responses
• Single-service deployment (frontend + backend together)
• Git LFS used for large ML model files
• Render-optimized production setup

---

🏗️ Tech Stack

Frontend

• React (Vite)
• JavaScript (ES6+)
• CSS (custom theme-based styling)

Backend

• Python
• Flask
• Flask-CORS
• Gunicorn

Machine Learning / NLP

• TensorFlow / Keras
• NLTK
• Scikit-learn
• Bag-of-Words model

Deployment

• Dockerized application (frontend + backend bundled into a single container)
• Render (Free Tier)
• GitHub
• Git Large File Storage (LFS)

---

📁 Project Structure

chatbot_project/
├── backend/
│   ├── app.py                 # Flask application entry point
│   ├── chat.py                # Chat response logic
│   ├── data_preprocessing.py  # NLP preprocessing (tokenization, stemming)
│   ├── init_model.py          # ML model loading & initialization
│   ├── intents.json           # Intent dataset
│   ├── model.weights.h5       # Trained model weights (tracked via Git LFS)
│   ├── data.pickle            # Vocabulary & labels (tracked via Git LFS)
│   ├── requirements.txt       # Backend & ML dependencies
│   ├── static/                # React (Vite) production build served by Flask
│   └── nltk_data/             # Runtime-downloaded NLTK data (ignored in Git)
│
├── frontend/
│   ├── src/                   # React source code
│   ├── dist/                  # Vite production build (copied to backend/static)
│   └── vite.config.js
│
├── Dockerfile                 # Single Dockerfile for full-stack deployment
├── .dockerignore              # Docker ignore rules
├── .gitattributes             # Git LFS configuration for large ML files
├── .gitignore
└── README.md