Professional
Send the footage you already record — broadcast, tactical camera, or phone — and every match comes back as movement your staff can search, compare, and question. Nothing new on matchday.
In practice
What the platform is built to change on the ground. The figures show example data, and every research or roadmap claim is labeled as one.
The weekend’s video goes in overnight. By the morning meeting, every player’s minutes exist as movement — sprint quality, stride symmetry, deceleration load — each read against that player’s own baseline, not a league average.
Where something drifted, the report flags it: a signal with the clips attached, for your staff to judge. Flagging drift early is a research program — the platform labels it as exactly that, and it never diagnoses.
Load planning starts from what the match actually took out of each player — not minutes played, but how movement changed while they played. The week takes shape from there: who needs volume, who needs recovery, who needs neither.
A squad-wide readiness view — intensity, coordination, and recovery markers in one place, before selection — is on the roadmap, and it is labeled Roadmap wherever it appears.
Scouting starts with a question your own archive can answer: who moves like the player you are about to lose? Movement search and player profiling rank candidates by how they actually move — and every name on the shortlist arrives with the supporting clips attached.
Profiles are built from movement, not appearance — and keeping them stable across footage sources, by re-identifying a player from movement style alone, is one of our research programs.
Capabilities
Everything on this page starts with pose estimation: skeletons for all twenty-two players, read straight from broadcast, tactical camera, or phone footage. Match video is crowded and cut — players cross, cameras switch — so skeleton repair & 2D→3D lifting fills occluded joints and lifts flat detections into 3D before anything downstream sees them. Movement signatures then turn those skeletons into per-player numbers that hold steady across matches and camera angles, and identity from movement is the research program working to keep every profile attached to the right player from movement style alone — no legible jersey number required.
On top of that base, movement search retrieves every recurrence of a pattern — a counter-press, a mistimed jump — from a season of footage in seconds, and player profiling ranks scouting candidates by movement similarity with the clips attached. At team level, tactical fingerprinting characterizes each opponent’s recurring patterns, injury-risk signals is the research program behind InjuryRadar — an early prompt for medical staff, never a diagnosis — and match readiness is on the roadmap: one squad-wide view of who is ready to play. Each capability page states its method, status, and limits.
Workflow
Nothing changes about how you film, and nothing new goes on the players. The exchange is simple:
Match or training video you already record — a broadcast feed, the tactical camera, or a phone from the stands.
Clean skeletons for every player, movement profiles that hold up across matches, and reports written for coaching, analysis, and medical staff.
Video, nothing else. No vests, no beacons, no sensors on players — matchday stays exactly as it is.
Next
Pilots open in small cohorts — coaches and performance staff first. Join the waitlist, or start the conversation directly.
