An idea, a deadline, and nothing in production.

There is a date that matters: investors, a pilot customer, a market window. The product exists as a document and nothing is in production. The risk is not writing the code. It is spending the runway building the wrong slice of it.

We name the riskiest assumption first, then build the vertical slice that tests it. The app deploys from day one, so “works on my machine” is never a status. Every Friday there is something you can click through, and what you learn from it decides the next week.

The surface is cut, the quality is not. Real authentication, a real schema, errors that do not crash on bad input. Minimum is a decision about scope, not a licence for shortcuts, because the point of an MVP is to grow if it works.

A production MVP your first users can use, with usage measurable from day one. The codebase, in your accounts. And a written list of what we deliberately did not build, so the next round of scope starts from a decision instead of an excavation.

If it works, it keeps going as a product build, with us or without us. The code does not care who ships the next feature, and that is by design.