GymIA

AI-assisted workout planning and gym client management for fitness professionals.

GymIA is a full-stack web application exploring how large language models (LLMs) can assist physical education professionals and personal trainers in creating, customizing, and managing personalized training plans. It combines client activity tracking, AI chat coaching sessions, structured workout plan generation via OpenAI, and interactive PDF/print export capabilities.

Disclaimer: GymIA is designed as an assistant tool for qualified physical education professionals to review and validate, not as an autonomous prescription or medical advice service.

---

Features

• Personal Trainer & Client Management: Secure JWT authentication for instructors, client profile tracking, anamnesis, and check-in history.
• Smart Inactivity Analytics: Dashboard tracking client activity trends, peak gym hours, and retention, categorizing inactive vs. critically inactive members.
• AI Workout Plan Generation: Tailors training plans based on client experience, goals, weekly frequency, and physical limitations using OpenAI fine-tuned models.
• Interactive Workout Plan Editor: Professionals can adjust exercises, sets, reps, intervals, and load dynamically before finalizing.
• AI Coach Chat Sessions: Integrated chat session management for instructors to converse with an AI fitness assistant directly inside the workspace.
• Export & Print Ready: Export workout sheets to PDF and print-friendly layouts for clients.

---

Screenshots

Dashboard & Plan Overview	Workout Plan & Client Management

---

Tech Stack & Architecture

GymIA/
├── backend/                  # Node.js + Express API
│   ├── prisma/               # Prisma schema & MySQL migrations
│   ├── src/
│   │   ├── app/
│   │   │   ├── controllers/  # Auth, Clients, WorkoutPlans, Dashboard, Chat
│   │   │   ├── routes/       # Express route definitions
│   │   │   ├── services/     # Business logic & OpenAI service
│   │   │   └── middlewares/  # JWT auth & error handling
│   │   └── index.ts          # Server entrypoint
│   ├── docker-compose.yml    # Local MySQL container setup
│   └── .env.example          # Backend environment template
│
└── frontend/                 # React + Vite Client
    ├── src/
    │   ├── components/       # Reusable UI components & layouts
    │   ├── contexts/         # Authentication & preferences context
    │   ├── pages/            # Dashboard, Clients, WorkoutPlan, Chat, Login
    │   ├── services/         # Axios API client & endpoints
    │   └── styles/           # styled-components global theme
    ├── public/images/        # Demo previews & brand assets
    └── .env.example          # Frontend environment template

---

Getting Started

Prerequisites

• Node.js (v18+) and npm
• Docker & Docker Compose (or a local MySQL 8 instance)
• OpenAI API Key

---

1. Database Setup (Docker)

Start the MySQL database container:

cd backend
docker compose up -d

---

2. Backend Setup

From the backend/ directory:

1. Copy the environment template and fill in your keys:

   cp .env.example .env

2. Configure .env:

   DATABASE_URL="mysql://root:password@localhost:3306/gym"
   JWT_SECRET="your-super-secret-jwt-key"
   OPENAI_API_KEY="sk-..."
   OPENAI_FINE_TUNED_MODEL="ft:gpt-4o-mini-2024-07-18:your-org:model"
   PORT=3000

3. Install dependencies and apply Prisma database migrations:

   npm install
   npx prisma generate
   npx prisma migrate deploy

4. Start the backend development server:

   npm run dev

   The API will be running at http://localhost:3000.

---

3. Frontend Setup

From the repository root, navigate to frontend/:

1. Copy the environment template:

   cd frontend
   cp .env.example .env

2. Configure .env (optional, defaults to http://localhost:3000):

   VITE_API_URL="http://localhost:3000"

3. Install dependencies and start the Vite development server:

   npm install
   npm run dev

   Open http://localhost:5173 in your browser.

---

API Endpoints Overview

Method	Endpoint	Description
POST	/api/auth/register	Register new instructor
POST	/api/auth/login	Authenticate instructor & return JWT
GET	/api/auth/me	Fetch authenticated user profile
GET	/api/clients	List instructor's clients
POST	/api/clients	Create new client
GET	/api/dashboard/statistics	Dashboard retention & inactivity statistics
GET	/api/workout-plans	List workout plans
POST	/api/workout-plans	Request AI-generated workout plan
POST	/api/chat-sessions	Create coaching chat session
POST	/api/chat-sessions/:id/messages	Send message in chat session

---

Security & Responsible Disclosure

• Never commit real credentials, API tokens, or production database strings.
• Always configure .env locally using the provided .env.example templates.
• Any generated workout recommendations must be evaluated and approved by a qualified personal trainer or fitness professional.

---

License

This project is licensed under the MIT License.

Developed by João Gabriel Alves Junior.