Vision
Beyond Football
The same movement intelligence applies wherever bodies move — running, dance, rehabilitation, fitness. “WithoutBall” becomes “Without[X]”.
Principle
One backbone, many heads.
The foundation model never learns “football”. It learns how joints move together — gait, load, symmetry, coordination. Verticals differ only in the lightweight heads and the product language on top; the representation underneath is shared.
How that split is engineered is laid out in the docs.
- Football
- Running & gait
- Dance
- Rehabilitation
- Fitness & longevity
Verticals
Where the same model goes next.
Football
PrototypeThe first vertical, and the proving ground: nine capabilities from pose estimation to match readiness, built on movement data derived from large volumes of match video. Everything the other verticals inherit is validated here first.
Running & gait
RoadmapStride mechanics, left–right asymmetry, and how both drift across a season of training. The same windows-and-embeddings machinery, pointed at a runner instead of a squad.
Dance
RoadmapMovement quality is the entire product here. Two concepts share the backbone: ChoreoLab for studios and companies, MirrorAI for personal practice — both at concept stage, both waiting on the same representation.
Rehabilitation
RoadmapMovement-pattern change over the course of recovery, measured the same way at every session. It supports the judgment of the professional in charge — it does not replace it.
Boundary: WithoutBall tracks movement patterns to support clinical decisions. It is not a medical device and does not provide a medical diagnosis.
Fitness & longevity
RoadmapHow you move is a long-horizon signal of how you are doing. A phone camera and a consistent measure of movement quality, year after year.
The name
“WithoutBall” becomes “Without[X]”.
The name states the method: take the ball away and the intelligence remains. Strip the sport away entirely and it still remains — a movement intelligence company that started with football.
WithoutBallWithout[X]