How we price a fixed-scope AI build

Most AI projects are sold by the hour and end without a finished product. We sell a plan and a price, agreed before we start.

Open-ended AI work has a familiar shape. A day rate, a vague scope, and a budget that grows every week while the thing you wanted drifts further away. The incentive is backwards: the longer it takes, the more it pays.

A plan and a price, up front

We scope one workflow worth automating, write down exactly what done means, and put a fixed price on it before you commit. You know the shape of the engagement on day one. If the scope changes, that is a conversation, not a surprise on an invoice.

Three phases

A build runs in three phases. Discovery finds the one workflow and the plan. Build ships it, with a working slice in week one and a demo every week after. Operate keeps it running once it is live.

Why fixed scope is possible with AI

The trick is to scope narrow and prototype first. We do not price a vague AI platform. We price a specific job the software has to do, and we prove the hard part early, so the estimate rests on something real instead of hope.

One part is not fixed, on purpose. Operating a live product is month to month, because a thing in production keeps changing and needs someone watching it. The build has an end. The product does not.

Fixed scope is not about being cheap. It is about removing the fear that you pay for months and end up with a demo. You get a plan, a price, and a product that ships.