Five steps, and what each one takes off the bill.

The shape of a month

You describe the business on a call. The transcript becomes requirements, a prototype and an architecture. Then a live board you can open whenever you want, weekly billing, no status meetings. At handover it’s documented and conventional enough that your own tools can extend it — and the documentation has three tests to pass before that counts as done.

Why the price works

• Direction, not requirements. You say what you’re thinking. I don’t ask how the software should work, because you don’t know yet and asking forces the decision early. Removes · the written brief, and the misreading that follows one
• Build to think. Rather than interview you into a document, the fleet builds something that embodies the idea and you react to working software. The building is the requirements process. Removes · the discovery phase, and its invoice
• Iterate until the real priorities surface. Each cycle moves something. Features that felt critical turn out to be later work, and things nobody could articulate turn out to be first. Removes · the feature built before the real one was known
• Find the invariant pattern. Under 457 job boards, or five CRMs, or every assistant’s private habits, there is one shape. Finding it is the architecture, and it is the part that isn’t delegated. Agents fill a shape in; they don’t find it. Removes · the bespoke one-off, and its maintenance bill
• The fleet builds against the pattern. Bare-metal Docker on commodity servers, conventional enough that the next person can read it. Removes · the five-person team, and the DevOps seat

Step four is the one that decides whether the rest works. It is also why the comparison on the price page is against one engineer rather than a team: one person finds the pattern, and the fleet’s throughput goes into filling it.

How the fleet is supervised

Nobody reviews the agents while they write. Given room, an agent invents work that doesn’t need doing, so I let the fleet run and bin whatever comes back wrong. What survives is architecture I’ll put my name on.

That is a real risk and it is worth naming: the review is at the output, not the keystroke. It is why the handover is rehearsed halfway through rather than promised for the end, and why the decision records are published.