Rebuilding the home base

I keep moving. Every move makes me rethink what "home" actually means.

The thing nobody tells you about a non-conventional path is that the path doesn't reveal itself as non-conventional in real time. You just keep making the next decision based on the question you can't stop asking, and then one day you look up and notice you're nowhere near where you were supposed to end up. Which is the point, I think. If you let other people draw the map for you, you end up living inside someone else's idea of a good life. I'd rather make the map wrong on my own terms.

Atlanta

I grew up in Atlanta. Humid summers, basketball on driveways, neighborhood kids who turned into family the way neighborhood kids do when there's nothing else to do but figure out what you love. I loved hoops. I loved making things. I was making basketball edits on social before I was writing real software. The two didn't yet know they were the same thing.

That's the kid I keep checking my work against. Would the version of me with a basketball and a notebook be excited by what I'm spending today on? It's a more useful filter than most of the ones the world hands you.

The leap

College was Berkeley. Three thousand miles from anyone I'd known. I picked it because I wanted my back against the wall, and the deep end of the pool is the only place I've ever actually learned to swim. CS and Finance, which sounded ambitious until I realized everyone around me sounded ambitious. I figured out pretty fast that the people I respected were less impressed by the labels and more impressed by what you'd actually shipped lately.

I shipped some things. I missed some things. I wrote my first real machine learning code. I started taking markets seriously. I started thinking about software less as a job to grow into and more as leverage to use on whatever question you couldn't stop chewing on.

Seattle

Amazon next. Seattle. ML and AI engineering at a scale I'd never seen and won't get to see again the same way. The thing nobody warns you about big tech is that the engineering rigor is real and the cadence is real. The other thing nobody warns you about is the way the vest works on your decision-making. Every quarter the next tranche gets closer. Every quarter the "just one more year" math gets easier to do. The pull isn't malicious. It's just the gravity of a system that's been tuned for a long time to keep good engineers in the chair.

I sat with that gravity for longer than I wanted to. I think about it a lot. How many of the smartest people I worked with have a startup idea they've been polishing in their head for three years and a vesting schedule that doesn't end for two more. How many of them will still be there in five years because the math, on paper, never quite tilts the other way.

I left with student loans still on the books and most of the salary I'd been earning about to evaporate. That part was uncomfortable in a way I don't get to talk about much because it's still recent enough that admitting it feels like admitting a risk I haven't fully cleared. But if I'm being honest, that's the chapter that taught me what I'm actually made of. You don't really know what you'll do under pressure until the pressure is your own bank account.

The pivot

Cerebro is where I left the warm part of the pool. Small team, big problem, sports tech. I led product, I led the AI side, I hired people who started as my friends and watched them turn into engineers. I learned what it feels like when there's no architecture committee and no review board. Just you, the customer, and whatever you decided that morning.

The lesson from that chapter that's stayed: the sports tech landscape has more white space than the headlines suggest. Most of what looks built isn't. Most of what's built isn't what teams actually need.

That second sentence is what made AirPLAi possible.

AirPLAi, with Danish

Co-founding AirPLAi Sports with Danish is the rebuild this post is named after. The website was just the easiest part to write about first.

Danish was a mentor first, then a friend, then a co-founder. The order matters. Most co-founder relationships I see start as friends who decide to build something together. Ours started as someone two steps ahead of me on the road, looking back, telling me what I was about to figure out the hard way. That kind of relationship makes the company feel less like a transaction and more like a continuation of an ongoing conversation.

The thesis underneath the company is one I'd been chewing on for years before I had the language for it. Vision models got good enough this decade that machines can finally watch sport the way the rest of us do. Multi-modal systems can read a court, follow a possession, name the action. The API cost curve is the only thing standing between today and a world where every game at every level, from a phone on a tripod at a rec gym to a broadcast feed at a pro arena, gets the same kind of intelligence the top of the pyramid takes for granted. That's a fight worth picking. That's the question I can't stop asking.

We're five humans now, plus a fleet of agents that build alongside us because the math of the market we're going after doesn't work any other way. The operating system runs the company. The humans handle the irreducible part. I write this because I'm not sure most people building software companies have noticed yet how much of the org chart can now be automated, and I'm pretty sure the ones who have noticed are the ones who'll get to choose what they build instead of inheriting it.

What I'm rebuilding

The "home base" of the title isn't really the website. It's the version of work I want to be doing five years from now. Closer to the things I cared about as a kid in Atlanta. Closer to the rigor I picked up at Amazon, without the wrong-target part. Closer to the small-team conviction I learned at Cerebro, with more leverage. Built alongside people I trust, on a stack I keep iterating on, with a story I'm willing to tell out loud.

Whether any of this lands the way I'm hoping, I genuinely don't know. But the moving is part of the work. I'd rather be moving than be sure.

What would you build if the vesting schedule wasn't doing the deciding for you?