How my AI tennis-coach sleeve placed 4th of 16 in a Physical AI comp

RYVO: a smart arm sleeve that buzzes the exact spot your swing went wrong. Feel the correction, not just read it.

Sensors on the sleeve track the arm through each stroke. A model on the phone ranks the single biggest correction, and a local safety layer decides whether to buzz. The sleeve then vibrates on the right spot, almost instantly. A cloud model (Gemini) writes the tip and a drill afterwards, since it is too slow for real-time feedback.

Tennis first, then basketball shooting form, then golf.

Business case. Paying customer, demand, defensibility, risks, and a data-handling policy for athletes' movement data.

Phone app prototype. Home, Train, Coach and Progress screens. A 3D replay of the arm with a reference overlay, a simple elbow cue, a focused drill, and a comparison across attempts.

Sensor bench. Bluetooth connection to the sleeve's five motion sensors, plus recording, review and export of each session.

Pitch Day scored the proposal: business potential 40%, technical feasibility 25%, innovation 25%, team and pitch 10%. We came out 2nd of 16 and received the finalist grant.

Demo Day scored the working product. We finished 4th of 16, as teams with stronger working hardware moved ahead.

A strong business case gets you into the final. At the final, the working product carries the score.

To be added.