Service · Scope

Find the problem. Then decide what AI has to do with it.

Most clients don't arrive with a clear problem statement; they arrive with a feeling. A one-week sprint names it, maps the top three friction points, and recommends a fix (AI, automation, neither, or both) with rough economics attached.

Fit

Who it's for, who it isn't

It fits when

  • You suspect "we should be doing something with AI" but can't name the what.
  • Your team is carrying operational drag that hasn't been explicitly measured.
  • You want an independent read before committing to a platform decision.

It doesn't fit when

  • You already have a defined use case and need delivery: go to Rapid Secure AI POC.
  • You need a compliance-specific review: go to AI Governance Assessment.
  • You want a workshop without a recommendation: that isn't us.
What you get

The deliverable

  • A short-form discovery report. Current state, top-3 friction map, candidate interventions per friction, indicative cost and effort, and a recommended first move.
  • A recommended next engagement, or a recommendation to stop and wait. We say both out loud.
  • A written summary you can take into a board meeting without translation.
How it works

The phases

  1. Day 1 · Kickoff + stakeholder interviews

    Meet the operational owners, the budget holder, and anyone who has to live with the outcome.

  2. Days 2–3 · Walkthroughs + desk research

    Observe the workflows that hurt. Read what you already have. Compare against what peers publish.

  3. Day 4 · Synthesis

    Pull the friction points into a map. Score them by impact and effort. Draft the recommendation.

  4. Day 5 · Readout + written report

    Joint session with the people who were interviewed. The report lands the same day.

Duration & price

What to budget

Duration

1 week

Price shape

Fixed fee

Fixed upfront, invoiced at kickoff. No scope creep, no day-rate surprises.

Sample deliverable

What the output looks like

An anonymised sample of a past deliverable for this engagement is being prepared. Until it's published here, the clearest picture comes from the methodology page. This service is one productised slice of the same method.

Summary

Most AI programmes fail because nobody stopped to ask what problem they were actually solving. This is the cheapest work we can do for you, and almost always the most valuable.
Not the right fit?

Try one of these instead