Capture
One working set. Full body in frame.
Record in the app or choose a clip you already filmed. Keep the set short — about 45 seconds, under 80 MB — so the lift is the lift, not a session montage.
Form before load
The shape is the work.
Record a set. Read what held, what to fix, and the next cue. Send a working set when you want a person, not only a measurement.
The read
Score, what held, what to fix, and the next cue.
UNITY measures the set and returns a form read: what you did well, the focus, and a cue. Replay can show the original clip or the tracked movement, including a measured bar path when the clip supports it.
Comparison
You against your best, your last time, or your coach.
Overlay or side by side. A coach can mark a reference, request a specific lift, agree or disagree with a finding, and leave a review note.
Privacy
The clip is not training data for a general model.
Movement media is stored privately for you and your assigned coach. Delete the analysis or the original video when you want it gone. UNITY Trainer still holds setup, execution, faults, and breathing for the movement in front of you.
Two doors
Measure it, or send it.
| Form intelligence | Form check | UNITY Trainer |
|---|---|---|
| A measured read of the set you just filmed: score, focus, cue, replay, and comparison. | One working set sent to your assigned coach, with a note about what to watch. They review the clip in the thread of the practice. | Setup, execution, the standard, faults, and breathing for the movement in front of you — before the load becomes the point. |
Replay can show the original clip or the tracked movement. When the clip supports it, the bar path is measured. When it does not, UNITY says the path is estimated.
Compare this set with your best, your last time, or a coach reference — overlay, or side by side. A coach can mark the reference, request a specific lift, agree, disagree, or override a finding.
You are asked before movement media is stored. Clips stay private to you and your assigned coach. Delete the analysis, or delete the original video. Re-measuring keeps the original result.