A hands-on diagnostic surface for serious AI-program operators. You submit a structured read on where your program sits; we return a Foundry-shaped diagnosis. The interface is the entry point to a real engagement.
The full surface (schemas, examples, runnable mocks) is shared under NDA. Self-selecting filter: the people who actually want to integrate against an interface are the people we want to talk to.
Most engagements start cold: a discovery call, a slide deck, an "intro to the practice." The API is the alternative. By the time we talk, you've integrated against the contract we work in and you know whether it fits.
What you get is plain enough: ladder position, weakest of the four basics, recommended engagement tier. What we don't put on the public web is the rubric behind it. That's the part that travels with us; that's the part that's gated.
If you'd like to integrate against the real surface, let's scope it.
Reach out →