Gymini

Inspiration

We wanted to make gym training more accessible and less confusing by combining AI with real-time fitness guidance. Gymini was inspired by the idea of giving users a personal coach directly on their iPhone.

What it does

Gymini can recognise gym equipment, provide personalised workout advice, track exercise form and repetitions using the camera, and give voice-based coaching feedback.

How we built it

We built Gymini as an iOS app using SwiftUI, Apple Vision for pose tracking, Apple Foundation Models for AI coaching, and ElevenLabs for voice output.

Challenges we ran into

Our main challenges were integrating multiple AI technologies, handling real-time camera and pose data, testing AI features on the simulator, and keeping the experience responsive.

Accomplishments that we're proud of

We successfully combined equipment recognition, real-time exercise tracking, personalised AI coaching, and voice feedback into one integrated fitness app.

What we learned

We learned how to work with Apple’s AI and Vision frameworks, integrate third-party AI services, and design an AI-powered experience around real user needs.

What's next for Gymini

Next, we want to support more exercises, improve form analysis accuracy, expand personalised workout planning, and make the coaching experience more adaptive to each user.