Blog · Studio notes
What we're
thinking
about.
Notes from the studio on AI-first engineering, design choices we keep making, and the work we do for clients. Long-form, opinionated, occasionally wrong.
More notes
Newest first.
What we actually use day to day.
An honest tooling rundown — which models we reach for, which we've quietly stopped using, and what surprised us. A frozen-in-time snapshot from the studio.
Your spreadsheet is already the app.
The pricing tab in your operations folder isn't a spreadsheet. It's an application — with hidden logic, fragile state, no audit trail. The cost of upgrading it just dropped.
The MVP died and nobody noticed.
"Minimum viable" was the right answer when building was expensive. When the real version takes a week instead of six months, the vocabulary starts distorting decisions.
A model is not a moat.
For business owners asking what their AI strategy should be — the durable advantage isn't which model you pick. It's the workflow you wrap around it.
AI capabilities don't really jump.
From outside, AI progress looks like a series of cliffs — one model can't do the thing, the next one can. From the inside, almost every capability worth caring about has been improving steadily the whole time.
A free first look at what your business could automate.
We're opening up a no-pressure first conversation — an hour with us, on the house, to walk through your business and point to the places where AI or automation could plausibly take work off your plate.
Custom apps cheap enough to throw away.
Software priced like a subscription assumes it'll live for years. Plenty of the most useful apps in your life would only need six weeks. AI finally makes that math work.
The maintenance was always the expensive part.
Custom software's real cost has never been the build. It's the years that come after — drift, scope creep, vendor lock-in. AI is the first thing that meaningfully changes that arithmetic.
Introducing AI-first development.
What it means to build software when AI is the default tool, not the novelty — and why the gap between what's possible and what most teams attempt keeps widening.