AI engineer. Product
Ships RAG, tools, evals, and guardrails into the real app. Notebooks are not the deliverable.
AI engineering is a practice we apply on product work. Agents, retrieval, evals, and mobile intelligence where the workflow needs them. Capacity, named people, and how they sit on an engagement are confirmed when work is scoped.
Ships RAG, tools, evals, and guardrails into the real app. Notebooks are not the deliverable.
Serving, cost, latency, and observability. The agent has to stay cheap enough to leave running.
KMP side intelligence. On device where it must be. On the server where it should be.
Architecture, the money cut, and the refusal. AI does not get to invent scope.
We do not publish a staff grid. Names and roles for an engagement are confirmed when work is agreed. This page describes the kind of AI engineering we discuss. It is not a headcount. It is not a hiring notice.