Ostra Bakery

The brief

Ostra came to me with a clear goal and a messy starting point. We agreed early that the work would be judged by what people did with it, not by how it looked in a deck.

What we made

I spent the first two weeks with the people who would use it every day, then cut the scope in half. What was left was small enough to get right: a few screens, a type system, and rules the client could keep using after I left.

Every decision went into a shared file with the reason next to it, so nothing depended on me remembering why.

What changed

Opened with a queue around the block. More useful than the number: the team now ships changes without waiting for a designer to approve them.