AI consulting

A board-ready view of where AI pays back in your operation, with the engineering cost attached.

Talk to an engineer

Typical duration: 4 to 8 weeks.

When to call us

For boards and leadership teams who need to decide where AI is worth the investment, with the engineering cost on the table.

  • The board is asking for an AI plan, and you want one that survives scrutiny.
  • Vendors are pitching, and you have no independent way to judge them.
  • Several departments have started their own experiments.
  • You suspect your data isn't ready, and want to know how far off it is.

What we deliver

Our consultants are the engineers who would build the thing, so every recommendation comes with a build estimate. Expect some of your ideas to come back marked not worth it.

  • A review of your processes, data and systems, done on site with your teams.
  • Use cases scored on payback, feasibility, data readiness and risk.
  • A security and compliance position your legal team can work with.
  • A build-or-buy recommendation for each use case.
  • A 12-month roadmap with budgets, written for the board.
Yours at handover
Assessment report, scored use-case register and roadmap.
Typical duration
4 to 8 weeks

How it moves through the shop

The same four stages as every project, applied to AI consulting.

  1. 1

    Assay

    This service is an assay: interviews, a review of systems and data, and a scored register of use cases.

  2. 2

    Forge

    Where it helps the decision, we build one small prototype on your data.

  3. 3

    Temper

    We test the recommendations with your security, legal and finance leads before they reach the board.

  4. 4

    Service

    A quarterly review of the roadmap, if you want one.

Read the full process

Questions about AI consulting

Are you tied to any vendor?

No. We don't resell software and we don't take referral fees.

What do we receive at the end?

An assessment report, a scored register of use cases, a build-or-buy recommendation for each, and a 12-month roadmap with budgets.

What if you conclude we shouldn't build anything yet?

Then the report says so, and sets out what would need to change first. That is a legitimate result of an assay.

Read all questions

Bring us the system you can't afford to get wrong.

Talk to an engineer

30 minutes. Bring the problem, and we'll tell you whether we're the right shop for it.