Jordan Lyall was sitting on his couch on a recent Saturday, trying to keep up with approximately 1 bajillion college football games across nearly that many streaming services, when he had an idea. Lyall pulled out his phone and began composing a message for Claude, the AI app from Anthropic.
I want to build an Apple TV app that shows what’s on across all the streaming services. Right now. Mostly interested in live sports, but I can see it growing to all content. I know Apple TV tries to work with apps but it doesn’t always work. I’d like to see just what sports are live now. And would be great to link or deep link into the app. Is this possible? Can we get the data? Does something like this exist? Would Apple approve it?
Two days (and countless messages with Lyall’s AI coding assistant) later, he posted the first screenshots of Clutch Time online. “I may be the only one who ever uses it, but it’s all I’ve ever wanted lol,” he wrote.
By Friday, 400 more people had signed up to use the app on Apple’s TestFlight platform.
“I’m totally geeking out about it,” Lyall said. “It’s super fun.”
— Jordan Lyall (@JordanLyall) October 5, 2026Over the weekend I built my ideal sports app
Every game, sorted by “best game on”. A heads-up when something rare is about to happen. Where to watch it. No ads. Apple TV too.
I may be the only one who ever uses it, but it’s all I've ever wanted lol
You can just build things pic.twitter.com/zkwai9qIal
As AI programming (AKA vibe coding) services have improved rapidly in recent months, fans are increasingly taking the tools into their own hands, creating fantasy trade recommenders, computer-voiced news digests, prop bet analyzers and even ticketing software. Across categories, the number of apps submitted to Apple’s store halfway through 2026 had doubled compared to the year prior.
Some apps are developed with only one person in mind. They’re built around features pro developers couldn’t monetize or needs so personal no one else would even think to meet them. Or maybe they’re just created to provide an ad-free alternative. Other creators set out with money in mind. All of it portends a world in which each sports fan can individualize their digital rooting lives without learning a line of code—if they care enough to tinker in the first place.
For now, the biggest public successes have come from those with at least some experience in the field. Lyall works as a chief product officer at a blockchain company. His development process turned out to be more complicated than it sounded at first.
While Claude Code is used to write the program, “sub-agents” are filling in gaps like identifying which channels carry each team’s games. Design-focused agents review each update and other bots suggest further improvements. Jason Weingardt, the developer of new MLB streaming app Mosaic, similarly had a background in product development when inspiration struck this spring.
“I remember having the thought this year of like, ‘Why do I need two devices to watch eight games on Opening Day?’” Weingardt said. So he asked an AI whether it would be possible to build an app with the ability to view many more games at once. Then he got more creative. “It started this sort-of waterfall of what-ifs,” Weingardt said.
He kept updating Mosaic daily, even as it stayed private due to potential licensing concerns. Weingardt would give his computer a task in the morning, monitoring its progress as he went about his day job. Then he’d test the day’s work during each evening’s MLB games. Eye-tracking on iPads proved capable of switching audio in multiview to the game someone was watching. Fans who signed up for the private app even got the ability to stream a game on their Apple Watch.
By July, the side project turned into a full-time job when MLB hired Weingardt as a principal product manager to explore changes to its flagship app. League engineers, like those across the industry, have been intrigued by the potential opportunities unlocked by AI-accelerated app development. MLB has a partnership with Google’s Gemini that includes fan engagement work.
“Everyone everywhere is using these things,” Weingardt said.
There are risks, of course. If every diehard is building his or her own apps, that’s support that’s no longer going toward the main apps used by mass audiences. Plus, AI users who go well beyond their coding depths may inadvertently unleash harmful bots on their own systems, their users or the leagues whose data they are after.
More customization and settings could come to existing apps, precluding the need for people to roll their own.
“Right now, it’s just for people like me who have built apps,” Lyall said. “I can imagine that, even apps where you’re locked into, they’ll make it even easier for you to just say a prompt, and you know, Instagram looks a little bit different. Or you know, my wife will be able to say, ‘I like the Facebook app, but I would love it if I could do this, this, and this,’ and then it’ll just work.”
Still, Weingardt would love to see sports properties allow their fans to continue tinkering, within reason.
“All it does is make people appreciate and consume more of whatever that game is,” he said. “As long as you’re entering into any of these projects with the right mindset … you know, building something cool for your friends and family, I think that that’s an easy thing for a league to endorse.”
But Weingardt also offered a bit of caution for any would-be fan developer: “Not everyone can go get their job because they tinkered in their basement.”