the shift

Before 2023, software was already the fastest way to make a difference in the world. But then it started to get even more accessible. Model capabilities convinced me engineering was about to become 100x easier.

In 2024, I started doing my day job at Microsoft in a quarter of the time, and spent all of my free time mastering what we now call AI engineering. I built almost all my ideas that were just sitting there. I quickly realized I wanted to make a living off of my own ideas, and began solving the puzzle of how to turn my small projects into a studio.

the studio

In early 2024, I joined Kinetic as an app studio: build our own ideas, launch fast, and see what worked.

As soon as we started hiring, I became lead engineer to run the team. The hardest part wasn't the code. It was finding good engineers who were actually using the tools of the times. Traditional hiring filters told us almost nothing about who could actually ship in this new environment.

In the first year, we shipped 12 apps. Three of them were successful, and we pivoted from an app studio to an incubator model, raising our Series A from GV.

focus

Eventually, two businesses were really scaling. That’s where I learned the hard lesson of focus.

The studio model had a built-in advantage: because there was always another project waiting in the backlog, we could kill ideas quickly without getting attached. For early exploration, that was the right call.

But when products have breakout traction, having an endless backlog of other things you could be doing becomes the enemy. Building something truly meaningful requires following a single thread for long enough to see where it goes. The company made the call to go all in on those winners.

what's next

Inspired by Mark Pincus's thesis in Life at the Speed of Play about separating your instinct from your idea, I knew it was time to start my own thing.

The timing was right to step out of the studio model, pick my own problem, and trust my own instinct on what comes next. More soon.