By Tanvi Bajaj
While bonding over their love for golf, Ivan Walls and Broderick “Brody” Higby (UC San Diego ‘19), noticed that many tools targeted advanced athletes, but barely any existed for those just looking to get into the sport. Inspired, they began developing GolfAI, an app that uses artificial intelligence to return recommendations and feedback to users after they upload a video of their golf swing. The central goal of the app, which the Triton entrepreneurs developed under their new startup called Data Hinge, is to improve athletics using computer vision.
Having an extensive background in computer science and machine learning, “I knew there had to be one thing I could do with computer vision to get better at golf,” Higby said.
We chatted with co-founders Walls, CEO, and Higby, CTO, about their GolfAI journey, along with their experience building a startup in San Diego’s growing tech ecosystem.
Year Founded: 2018
Key Players: Ivan Walls (CEO), Brody Higby (CTO), Natalie Karapetian
(Director of Operations)
Headcount: Seven interns, along with the leadership team
Headquarters: Qualcomm Innovation Space (Atkinson Hall, UCSD)
Technology: What makes GolfAI’s technology unique is the startup’s innovative software that uses advanced video analysis to improve your golf swing using only your smartphone’s camera. Rather than waiting days for feedback sent by a real golf coach, GolfAI simply provides real time, instant voice feedback after each swing via its built-in personal golf coach.
The technology is premised upon a predetermined big data set of major factors that every golfer shares (e.g. shoulder turn, hip turn, shoulder tilt, swing plane, ball flight). The system has an inputted ideal and bases its recommendations on that; but instead of linear information, it provides feedback on the most important things that can be fixed. Higby sums it up perfectly: “GolfAI uses machine learning to make learning golf easier and more efficient.”
Inspiration: Walls explained that the idea of creating a smartphone app came from the founders’ goal to find a clever way to make use of technology that people carry everywhere. “Users can monitor their swing no matter where they are, and as they improve, the data-driven technology is able to adapt and grow along with them,” he said.
Challenges: The biggest challenge for GolfAI’s founders has been a combination of refining their vision for the app as well as building the technology itself. Higby added that they had to develop 3D pose estimation, object detection for the ball and club, and the machine learning model for the predictor of a good golfer versus a bad golfer, as well as what part of the swing the golfer was in.
Higby described the process, saying “I built it using a gradient adversarial network,” afeat that took him about three months while he was still in school.
Funding: While GolfAI was initially self-funded (the founders contributed about $23,000), Higby and Walls said that winning Veteran Ventures and being finalists at Ignite (two competitions out of UCSD), gave the startup a big boost financially, and has raised the company’s profile.
Mentors: GolfAI’s current fundraising techniques are supported by the startup’s informal Board of Advisors, which includes Grant Olsen, Shane Loman, and Eric Gasser – all of whom Walls and Higby credit with helping them make connections, handle their finances wisely, and more.
On the Horizon: Over the next few years, Walls and Higby hope to refine their technology and expand it into different sports. Past that, they hope to look to different industries (such as healthcare) in and outside of the United States. In fact, Higby and Walls mentioned how UCSD’s School of Medicine has been interested in using the startup’s computer vision technology to detect the early onset of Parkinson’s. Higby explains: “It allows early detection of Parkinson’s by looking at the person’s gait or the way they walk. It’s not as noticeable in early-onset, which is where the software would help.”
The GolfAI team
San Diego Tech Ecosystem: Both Walls and Higby agree San Diego is the perfect place for them to headquarter their startup. Not only have they had a lot of interest here due to golf’s popularity, but also the growing tech scene makes San Diego a prime location for investment and development. Higby concludes that “the weather and culture lend themselves to a great place for a sports startup.”
Tacos: Walls’ favorite local taco spot is Taqueria Revolución in Westfield National City, whose tacos, he said, are as authentic as it gets. Meanwhile, Higby, who is vegan, prefers Tocaya in La Jolla.
Keep up with GolfAI through Instagram, Facebook and LinkedIn.
Editor’s Note: TritonTech is an original series on UC San Diego created by Fresh Brewed Tech that showcases the innovative ideas born in the halls of academia that are making a great impact on our ecosystem and beyond.
