Stepping up to the plate to simplify sports coaching

Computer science students are stereotyped as the polar opposite of sporty. Meet the student who’s colliding the two worlds — coaching and AI — together.

Grace Cowger | Graphics Editor

CoachVision.ai was brought about by computer science sophomore Nikhil Marisetty with the aim to help people who are getting into a sport for the first time.

The AI tool allows users to set up their phone and record themselves performing a particular movement in a sport. The app analyzes specific movements, especially the joints and how they move, eventually giving users tips on technique and form. Marisetty finds that these tips don’t necessarily need to have such a large price tag on them, since they can be learned easily with the app.

“Whether it’s private tennis lessons or private soccer lessons, I think especially in the U.S., it is very expensive these days,” Marisetty said. “I know people that pay like $100 an hour for lessons, [and] when you are just starting out and you’re just trying to learn the skills, I don’t think you should be paying that much to learn these skills.”

Marisetty also found a personal connection to his app, given he has experienced the need to get better at the sports he played for fun but didn’t want to pay the hefty price tag associated with professional instruction. What was a universal issue faced by many entering a sport no longer has to be, he said.

“I used to play a lot of sports with my friends and every time I thought my friends were better than me, I [wanted] to train in private,” Marisetty said. “But I also didn’t want to pay for coaching, so that was one of the motivating factors.”

Realizing this, Marisetty said his motivation to create the app only accelerated when he found himself playing soccer in his backyard.

“I love soccer, [but] I have never really trained,” Marisetty said. “I’d go outside and learn to kick the ball, and for some reason I could just never really lift the ball, [and] it would have no air time. Initially, I didn’t know how to place my foot or how to really shoot the soccer ball. In cases like that, the app can tell you whatever mistake you are doing.”

While the idea may be appealing for the novice entering a sport, coaches may be hesitant to get on board with the idea. Marisetty said his idea incentivizes a different approach to athletic instruction. In a world with accessible technical coaching from digital assistants, human coaches will focus more on the emotional connections to performance, something that can’t be replaced. Marisetty’s focus is to eliminate the large cost of entry to a sport instead of coaches’ jobs.

“I am not trying to put coaches out of jobs, but I think the cost of entry should be reduced,” Marisetty said. “[Take] soccer, at a certain point you know how to shoot or pass a ball is not enough. You need to be able to play with the team, you need to be able to learn tactics [and] strategy. That’s something our app can’t do and we’re not targeting that section of the market either.”

Athletes in sports have two focuses: physical and mental health. While the app currently focuses on the former, building technique that focuses on correct form, the mental health aspect is one that the app is yet to cover. Marisetty said that with the unlimited amount of possibilities AI has, mental health access is within an arms’ reach.

“If you’ve used ChatGPT before, even if your idea is pretty bad or what you’re saying is pretty bad, it’ll be like, ‘That is a great question,’ and ‘That’s a great idea,’” Marisetty said. “If you can just add a feature to the AI model, [where the user] wants a very encouraging coach, or a very strict coach, we could definitely add [that] feature.”

Sports has an element of trust between the athlete and the coach. While Marisetty’s idea doesn’t have a physical coach, he said that his app’s methodical approach to data collection is what makes it stand out from other AI platforms that exist.

Marisetty said that getting the app to reliably analyze user footage started with filming videos from sports professionals of various ages and levels of competition. Next, Marisetty approached coaches and asked them to provide feedback on the videos recorded. That feedback is then used to train the AI powering the app on how to analyze user videos.

“You can upload your video to ChatGPT and ask it for feedback,” Marisetty said. “But you’re not going to trust what it gives you. To be able to provide accurate feedback, we need a lot of data.”

Any athletic technique can have multiple interpretations from coaches, with each providing their method or ideas on how to successfully accomplish a task. CoachVision.ai plans to keep that concept alive, as Marisetty himself has seen how different coaches’ advice varies.

The app currently is limited geographically to the Dallas region, a purposeful decision by Marisetty. Since Dallas offers an expansive sports culture, he finds the city to be a good spot for incubation.

“I’d like to see people in the Dallas region using it,” Marisetty said. “I think this app could be very successful especially in Texas because of the large fields and all the play areas and there is a very big sports culture [here] as well. Beyond that, we can scale even higher. You never know where it can go.”

With entrepreneurship, there are skills a founder brings to the table, and those that they must learn to grow. Marisetty said it is his adaptability that has allowed him to see where this app has

its faults and how to improve those areas. Uniquely, he finds that through the process of being an entrepreneur, his personal skills too have grown, including his networking, public speaking and comfort in new situations.

“In entrepreneurship, you do have to give up because sometimes it’s not the right idea or you’re not doing it right in the right way, or at least change things,” Marisetty said. “I think I am adaptable in the way I can see where things aren’t going right, I just need to figure out the best way that users will actually use this.”

Marisetty won the UTD Founders Award and has seen a positive response to his work so far. While this app is in the very early stages of growth, he said it can be scaled up and built further after he completes the data collection.

“I know this market exists,” Marisetty said. “[Users can find] whether they really like a sport or not, if they realize they don’t like [the sport] as much they can switch to another sport, [and] they haven’t spent a lot of money trying it. The biggest thing you can get from this is finding out your passion for which sport.”

Leave a Reply

Discover more from The Retrograde

Subscribe now to keep reading and get access to the full archive.

Continue reading