TechBOT - IT Education Chatbot

A chatbot that provides IT learning resources, course recommendations, interview tips, and more using LSTM, FastAPI, and a frontend built with HTML & CSS.

🚀 Project Overview

TechBOT is an AI-powered chatbot designed to assist users in the IT industry by answering queries about programming, certifications, job opportunities, and interview tips. It uses LSTM for intent classification, FastAPI for backend deployment, and an interactive HTML/CSS frontend for user interaction.

---

📌 Features

✅ IT Course Recommendations - Suggests IT courses & certifications.
✅ Programming Guidance - Tips to learn Python, coding, and improving programming skills.
✅ Job Opportunities - Provides insights into high-demand IT jobs.
✅ Interview Preparation - Helps with coding interview strategies and system design concepts.
✅ LSTM-based Intent Recognition - Uses deep learning (LSTM) for chatbot responses.
✅ FastAPI Backend - API-driven chatbot with real-time responses.
✅ Frontend with HTML/CSS - Interactive web-based chatbot UI.

---

🛠️ Tech Stack

Backend

• Python 3.12.2
• TensorFlow/Keras (LSTM model)
• FastAPI
• Joblib (for saving tokenizer & label encoder)

Frontend

• HTML, CSS
• JavaScript (to communicate with FastAPI backend via AJAX)

Deployment

• FastAPI Server (Backend API)
• HTML/CSS Frontend (User Interface)

---

🛠️ Installation & Setup

1️⃣ Clone the Repository

git clone https://github.com/Ananthakrishnan12/TechBOT.git
cd TechBOT

2️⃣ Create a Virtual Environment & Install Dependencies

python -m venv Techbot
source Techbot/bin/activate  # For Linux/Mac
Techbot\Scripts\activate     # For Windows
pip install -r requirements.txt

3️⃣ Train the LSTM Model & Save Artifacts

python train_model.py  # Train the chatbot model

This will generate:
✅ chatbot.h5 - Trained LSTM model
✅ tokenizer.joblib - Tokenizer for text preprocessing
✅ label_encoder.joblib - Label encoder for intent classification

---

🚀 Run the Chatbot API

uvicorn main:app --reload

The FastAPI server will start at http://127.0.0.1:8000/

✅ Visit http://127.0.0.1:8000/docs for API documentation

---

DEMO :