AI GymBro is an AI-powered personal trainer that uses your device’s camera to analyze your squat and pushup form in real time. It provides instant, accurate, and encouraging voice feedback using advanced pose detection, LLM-based coaching, and realistic text-to-speech.

• Real-Time Joint Detection: Uses computer vision to track your body joints live.
• Exercise Recognition: Detects and analyzes squats and pushups automatically.
• AI Voice Feedback: Provides instant, personalized coaching using LLM and TTS.
• Rep Counting: Accurately counts your repetitions and gives motivational praise.
• Modern UI: Clean, responsive interface for both desktop and mobile.

• Frontend: React (Next.js), TypeScript, Tailwind CSS
• Backend: FastAPI, Python
• AI/ML: Pose Estimation (MediaPipe), LLM (Ollama/Mistral), ElevenLabs TTS
• Deployment: Docker, Uvicorn

• Node.js & npm
• Python 3.10+
• Ollama (for local LLM)
• ElevenLabs API Key
• (Optional) Docker

1. Clone the repository:

   git clone https://github.com/your-org/ai-gymbro.git
   cd ai-gymbro

2. Frontend Setup:

   cd app
   npm install
   npm run dev

3. Backend Setup:

   cd ../
   pip install -r requirements.txt
   python -m uvicorn ai_server:app --host 0.0.0.0 --port 8000

1. Open http://localhost:3000 in your browser.
2. Allow camera access.
3. Select your exercise (squat or pushup).
4. Start moving! Get instant AI voice feedback on your form and reps.

This project is licensed under the MIT License.

• ElevenLabs
• MediaPipe
• Ollama