Every product here began as a scattered idea, tested cheaply and assembled deliberately into something whole.
A few people set direction and make every call that matters, while an AI workforce does the heavy lifting around them. The studio runs the same way its products work, by design.
Specialist agents carry the volume of the day-to-day: the research, the marketing, the books, and a good share of the engineering.
Agents prepare the work and people make the calls: every gate on the pipeline, whether greenlight, launch, or kill, ends in a human decision made on the record.
Headcount stays small because the leverage comes from the workforce and the shared studio roof, both of which scale long before hiring would need to.
What ships is the same bet the studio runs on: software that puts AI to work on one real problem, operated as a business of its own.
Five stages between a note in the backlog and a running business.
Ideas enter a shared backlog and get scored against the studio's thesis. Most stay there by design, because the filtering starts long before any money moves.
Before a line of code, every candidate faces the greenlight gate: demand, economics, and fit with the studio's strengths, checked in writing.
One product at a time, taken to 100% readiness. Core functionality, pricing and billing, and the marketing site all ship together as one finished release.
The same launch playbook runs every time: the product, the pricing, and the story go out together, so day-one users get the finished thing rather than a rolling beta.
Each launch joins the portfolio with its own brand and P&L, while the studio carries the shared roof and retires what stops earning its place.
Today, a product studio: we build and operate our own AI-powered products under one roof, one P&L per product. The ambition is a venture studio: products that keep earning their place graduate into companies of their own, with the studio as their first backer. For now we use the smaller word, and we are building toward the bigger one.
No. We only build and operate our own portfolio. If you want to talk about a partnership or something else, get in touch.
There is a kill switch at every stage, launch and operate included. Every kill produces a memo: the thesis, the test, the numbers against the bar, and what gets salvaged. Killed ideas are archived rather than deleted, so the next similar idea starts from that memo instead of from zero.

Notes on tech, startups, and AI, plus the odd bit of studio news, arriving only when there is something worth saying.