Transforming Basketball with AI
Today, on behalf of NEX Team, I’m thrilled to share the news of our latest endeavor, HomeCourt — the first AI-powered basketball app that uses the latest mobile AI to record, track, and chart shots in real time for players of all levels. Our launch also coincides with the completion of a $4 million seed round. We’re excited to have investors who believe in our mission to transform basketball with computer vision technology, including Mark Cuban, Steve Nash, Jeremy Lin, Sam Hinkie, Mandra Capital, and Charmides Capital.
The inspiration for HomeCourt came one evening last year after a weekly pickup game with friends. My wife and daughter had stopped by to cheer me on, but I couldn’t make a single shot with them watching. I thought: I wish they could see a short highlight reel of how I normally play when they’re not around. That spurred a ton of other ideas around performance tracking and player development that all centered around the desire to do all this using just my phone.
I had already left Apple by then, and it wasn’t long before a core group of us who’d worked together before started whiteboarding, breaking down the question of how to transform our phones into a training companion for basketball. We were excited about the problems we could solve with AI technologies, but by this time last year, they weren’t quite far along enough for what we had in mind. However, these technologies were evolving rapidly.
The Last Year
Last May, reliable object detection frameworks had just become available on mobile devices. A month later, at WWDC, Apple announced CoreML. Big improvements in other key AI technologies like Pose Estimation started to become available, which inspired us to develop pose estimation technology on mobile devices. And by last summer, we had enough to start putting the puzzle pieces together.
Our team got scrappy and started working with the technologies that were immediately available, but advanced them as well (growing our IP portfolio along the way). Within a few months, we’d made enough breakthroughs that Apple started touting HomeCourt as an app that exemplifies the power of CoreML.
By February of 2018, we had enough of a beta app to launch an early access program at the MIT Sloan Sports Analytics Conference. Since then, players and teams worldwide have actively used our app for training and development, including the Sixers and Celtics and top D1 programs like Duke University, Florida University, and Stanford University.
But HomeCourt isn’t just for professional athletes — it’s built with everyone in mind.
For instance, high school teams like the Mitty Monarchs are using HomeCourt. And in just five months, our active player base has spread to all 50 states, over 50 countries, and over 1,000 cities.
More to Come
Today, we’re finally launching HomeCourt out of beta, but this is just the start. We already know that the technologies we’re building can do far more than the tracking that HomeCourt provides today. On the player development front, we’re already training AI to extract much more information from video footage, including categorization of shot types, release time, speed of movement, and jump height — all in real time. We’re also working hard to track not just your individual practice session, but sessions with others too like a 1-on-1 game or a 2-player workout. But even these features just scratch the surface.
Our vision for HomeCourt extends far beyond its practical aspects as a training companion. By combining the latest in mobile, AI, and computer vision technologies with deep community engagement experiences, we believe HomeCourt can connect a worldwide community of athletes who share a love and passion for basketball.
So whether you’re just starting out like my daughter, Chloe, or playing at the collegiate level, we hope you’ll give HomeCourt a try!
To get started, just download HomeCourt from the App Store on your iPhone or iPad.
