ASSESS

Senior eyes on what you have, before you scale it.

A fixed-scope review of an AI system, an engineering organization's way of building, or a platform you intend to build on. You get written findings, in priority order, from people who have run these systems at enterprise scale.

THE WORK

Three reviews we run.

  • 01

    AI readiness review.

    Where the organization stands on data, platforms, skills and governance, and what has to be true before AI investment pays back. For leadership teams deciding how much to commit and in what order.

  • 02

    Architecture review.

    An AI system in build or in production, examined end to end: model selection, retrieval, prompts, evaluation, observability, cost, security posture. We write up where it will break at ten times the load and what to fix first.

  • 03

    AI-SDLC baseline.

    How your engineering team builds today, measured - cycle time, review load, test coverage, where AI tools are used and where they sit idle. The baseline that an enablement program is judged against.

DELIVERABLES

What you get.

  • A written assessment with findings ranked by risk and effort, and a recommended order of work.
  • A readout session with your technical leadership, with time for the questions the document raises.
  • Where it applies, a reference architecture or target-state sketch to build toward.

ENGAGEMENT

How it runs.

Two to four weeks. We read what exists, interview the people who built and run it, look at the code and the data, and write. One or two of your engineers spend a few hours a week with us; nobody is pulled off their work.

EXAMPLES

Typical scope.

  • A customer-facing AI assistant that works in the pilot and worries the team ahead of general release.
  • A platform decision - vendor, model provider, orchestration layer - that the company will live with for years.
  • An engineering organization of a few hundred people that wants to know whether AI tooling is paying for itself.

BOUNDARIES

What it is not.

Assess is a review, not a build. If the findings call for work, that is a separate conversation, and you are free to have it with anyone.

Next
Plan
Related reading
The AI-Native Operating System

Where is your AI work stuck?

First calls run 30 minutes over video. We will ask what you are building, what is in the way, and what a good outcome looks like - and tell you plainly whether we are the right people for it.

Contact us