Students Build AI That Spots Para-Athletes. Scouts Will Be Unnecessary Eventually. The Algorithm Is Not The Hero Here.
Samsung's Solve for Tomorrow program produced finalists including ParaSync, an AI platform that assesses adaptive sporting techniques and connects promising para-athletes with professional coaching. A separate team in Chennai is developing Rogue, a wearable for Parkinson's disease exercise support, and a Hyderabad student built Somatrace, an IoT wearable for movement tracking. These are students aged 20 and 21 building tools that most sports federations have not bothered to attempt.
This demonstrates what I call Distributed Talent Identification. The mechanism is that traditional scouting concentrates where infrastructure already exists, creating a sampling bias. AI assessment tools remove the geographic constraint by evaluating technique remotely, which means the pool of discoverable athletes expands without building a single new training facility. Talent was never concentrated in established sporting centers. Detection was.
Manoj Kumar Kasinathan, Jena R, and Chriswin William D, all around 20-21, are developing Rogue in Chennai. Homanshika Reddy Pullagurla, 21, develops Somatrace in Hyderabad. The ParaSync team are finalists in Samsung Solve for Tomorrow.
- Open any free AI video analysis tool, such as Google's Teachable Machine at teachablemachine.withgoogle.com, and record yourself performing a basic exercise like a squat.
- Upload ten short clips of correct and incorrect form as training data, then test the model on a new clip.
- Notice how the model identifies form differences. That is the same principle ParaSync uses to assess adaptive athletic technique, just at a dramatically smaller scale.