JPT — Your AI Personal Trainer

A full-stack AI chat application powered by GPT-4o Mini, built with FastAPI and React.

GitHub: https://github.com/Naknik3/jpt

---

Overview

JPT lets you have a real-time conversation with an AI personal trainer. Each chat session is stored in MongoDB so the AI maintains full memory of your conversation history, giving you coherent, context-aware responses.

---

Tech Stack

Layer	Technology
Frontend	React 18, TypeScript, Redux Toolkit, Vite
Backend	Python, FastAPI, Pydantic
AI	OpenAI GPT-4o Mini
Database	MongoDB (async via Motor)

---

Project Structure

jpt/
├── backend/
│   └── src/
│       ├── controllers/   # Route handlers
│       ├── services/      # Business logic & OpenAI integration
│       ├── models/        # Pydantic request/response models
│       └── utils/         # Config, DAL (MongoDB)
├── frontend/
│   └── src/
│       ├── Components/    # Layout, Header, Footer, Menu
│       ├── Screens/       # Jpt (chat), About, NotFound
│       ├── Services/      # Axios API calls
│       ├── store/         # Redux slices
│       └── utils/         # App config
└── database/

---

Getting Started

Backend

cd backend
python -m venv env
source env/bin/activate      # Windows: env\Scripts\activate
pip install -r requierments.txt

Create a .env file in backend/:

DATABASE_URL=mongodb://localhost:27017/jpt
OPENAI_KEY=your_openai_api_key

uvicorn src.app:app --reload --port 3000

Frontend

cd frontend
npm install
npm run dev

The app runs at http://localhost:5173.

---

API Endpoints

Method	Endpoint	Description
POST	/chat	Send a message and get an AI response
GET	/conversations/:id	Fetch conversation history
GET	/health	Health check

---

Features

• Persistent conversation memory across the session
• New chat button to start a fresh conversation
• Smooth auto-scroll to latest message
• Loading indicator while awaiting AI response
• Security headers + CORS protection
• Input validation on both frontend and backend