Writing
Thoughts on building, infrastructure, AI systems, and the people side of engineering.
Never Burn Bridges
Five years at Oscilar. I got hired because of a hack-a-month, an angel check, and friendships I kept for years — and that's the whole point: it's not your resume, it's what you do day in and day out.
I Wanted a Number, Not a Paragraph
When I built OfferPro, half of what I called 'AI' was just parsing the model's paragraphs back into numbers. So I test-drove TypeSafe's Jev — a model that skips the prose and returns typed decisions with confidence scores — by rebuilding OfferPro's offer-scoring in ten minutes.
We're Past AI That Answers Questions
Last week I registered my kids for a national chess tournament without opening a browser. We spent two years amazed that AI could answer questions. The interesting part now is AI that does the work.
Buy Next to Good Neighbors
We bought a house that looked like a downgrade on paper — old toilets, beaten floors, a murky koi pond. The bank made us write a letter explaining why. Two years later, the pond has baby koi and the neighborhood has our kids' best friends.
I Built the Tool We Wished Existed
Two years ago we were those parents — driving to tournaments every weekend with no real idea if any of it was working. So I built CoachMyChess. The AI was the easy part.
The Product-Minded Engineer
I've been the first engineer on a team more than once. In every one of those seats, the same trait mattered more than raw coding ability: caring about why you're building the thing as much as how.
What Nobody Tells You About Building Solo
People think the hard part of building solo is the code. It isn't. The hard part is that you are everyone — PM, designer, engineer, and the person answering the support email at 11pm.
Real Estate Tech: The Domain Beats the Stack
If you want to build real estate tech, your edge is not your stack. It's whether you actually understand the world you're building for. I learned this the expensive way with BetterCapital.
Knowing When to Step Away
We spend a lot of energy celebrating the decision to start something. We say almost nothing about the decision to stop. I built BetterCapital for five years — and then I wound it down.