Use cases

Every player leaves with their own clips

A tournament is filmed for eight hours across six cameras, and almost everyone in it goes home with nothing. Highlight Forge finds each player's moments while the round is still running.

Sector
Sports broadcast
Work
Product, AI, build
In production
A professional golf tour
The desk showing every live feed against one detection timeline
Every feed, against one timeline.
The attribution queue, showing the evidence behind each clip
Each clip carries the evidence for who is in it.
The clip editor with a cut moment and its metadata
A cut moment, with its player, hole and shot already on it.
Per athlete clip libraries with a reel ready to export
Every athlete has a library by the end of the round.
01 / 04
Challenge

Six cameras run for eight hours. One tournament produced 10,507 moments worth keeping, and a team watching in real time can cut the best few before the day ends. The rest are never named, so nobody can find them again, and the players in them go home with nothing.

Approach

The system watches all six feeds at once and works out who is in each moment the way a spectator would: the scoreboard, the kit, where the player is on the course, and what the commentator just said. It never reads faces. When two signals disagree, a person decides.

Outcome

5,519 clips came out of that tournament with a player, a hole and a shot already on them. The athlete, the coach and the sponsor get the moment the same day, and the editor picks instead of hunting.

Player names on these screens are invented. The figures are from one tournament in production.

Contact

Tell us the challenge.