In 2012, a high jumper described a problem to Patrik Olsson and the group that would start Spiideo. When he trained without someone to film him, he propped up a phone. The camera stayed still while he moved. Olsson, who had led software work at Sony, saw a problem software could help solve.
“a camera that could be mounted and autonomously track and film the athlete’s movements and technique”
A second conversation changed the scope
In Olsson’s account, the first idea was a mounted camera that could track an athlete through a training session. Then he and his colleagues spoke to a coach at Malmö FF. That conversation helped them see that teams also needed to record and review movement across a much larger space. Olsson and Örjan Johansson co-founded Spiideo in Malmö.
Giving the coach the film
Following a high jumper and following a team are different camera problems. They share a practical constraint: a coach or athlete who is busy with training may have no one available to film it. Olsson’s starting point stayed useful as Spiideo moved into basketball and other team sports. The camera would have to do enough by itself to leave people free to run the session.
AI in the product · November 13, 2026
Spiideo began with a camera that could keep up with an athlete training alone. In May 2026 it took on a job that comes after recording: turning game footage into highlights and short stories. The feature launched inside Spiideo Play and supports basketball.
Olsson’s team built AI Highlights with the AWS Generative AI Innovation Center. Spiideo identifies Amazon Bedrock foundation models and AWS Step Functions as parts of the system. That is a separate job from the sport-specific AutoFollow technology that keeps play in frame during a game.