Fourteen days from "we should do something with AI" to a decision you can fund.

• You have a process that looks like an AI problem, and no way to tell whether it is
• You have been quoted a six-figure build and cannot justify it without evidence
• Your team has run experiments that worked in a notebook and nowhere else
• You need something concrete to take to a board that has heard enough about AI

Week one — we find the real problem. We map the workflow end to end, sit with the people who actually run it, and assess whether your data is in a state that supports what you want to do. Most of the value is here, and most of the surprises are too.

Week two — we build enough to be sure. We design the target architecture, build a clickable prototype on your own use case, and estimate the cost for the full implementation. Then we tell you what we would do and what we would not.

Your team's total time commitment is a few hours across the two weeks. That is deliberate.

• A clickable prototype built on your own use case, not a generic demo
• A technical specification and target architecture your engineers can build against
• An implementation roadmap with phases, dependencies and a line-item budget
• A data readiness assessment — what you have, what is missing, what it will cost to fix
• A written recommendation, including the option not to build

Everything is yours. If you take it to another engineering partner, it will work there too.

• Not just a strategy deck. Nothing we hand over stops at the level of "consider adopting AI".
• Not a workshop. We do the work; you review it.
• Not an open-ended R&D budget. Fourteen days, fixed price, fixed scope.
• Not a sales process in disguise. If the honest answer is that you should not build this, that is what the recommendation will say.

Fixed price, fixed scope, two weeks. Larger or multi-system scopes are quoted after the scoping call.