TrackFit Rep Coach
Project Gallery
My role Product Designer · Hands-on Builder
Overview
TrackFit was built as an ambient bodyweight training product: a room-aware coach that can sit on a tablet, detect when someone is ready to train, and turn short rep sets into a more game-like routine.
Key features include:
- Pose-guided workout flows for squats, press-ups, and lunges with live rep counting and form feedback
- Always-on standby mode that uses room presence to wake a coach-like interaction loop instead of a standard app start screen
- Voice coaching and speech-triggered session control, paired with configurable coach tone and voice profiles
- Persistent player profile, streak tracking, daily and weekly goals, and lightweight progression mechanics
- Cloudflare-backed leaderboard and voice-worker services, with Capacitor packaging for mobile deployment paths
The result is closer to a productized training surface than a basic fitness tracker: computer vision, habit loops, and ambient coaching wrapped into one focused interface.
Key decisions
- Designed the experience as an ambient training surface rather than a conventional tap-first fitness tracker.
- Combined computer vision, voice, habit loops, and mobile packaging around short repeatable training sessions.
Evidence
- Live pose-guided rep counting and form feedback
- Room-presence wake flow and voice-controlled coaching
- Progression, goals and Cloudflare-backed leaderboard services
Technologies Used
- React
- TypeScript
- Vite
- MediaPipe
- Capacitor
- Cloudflare Workers
- Cloudflare D1
TrackFit is shown through product screens because the current app preview is local-only and there is not yet a stable public frontend URL to share.
