Three ways to start. One steady weekly rhythm.

Find where AI pays off, prove it on your data in four weeks, then roll it out.

01 · Our approach

Our approach in 4 steps

Find where AI pays off, prove it on your real data, roll it out, then make it part of how your team works. We stay until it is used in the real work, not just shown in a demo.

  1. 2 weeks

    Step 1: Find where AI pays off

    AI Opportunity Audit

    We talk to the people doing the work and rank where AI would genuinely help, before anything is built.

    • Interviews with the people doing the work
    • Use cases ranked by value and effort
    • A check of your data and systems
  2. 4 weeks

    Step 2: Prove it on real data

    4-Week Working Proof

    We build one use case on your real data and tools, then pilot it with the people who will use it.

    • A working build connected to your systems
    • Weekly demos with your team
    • A clear go / no-go against agreed measures
  3. After the proof

    Step 3: Roll out what worked

    Focused build or partnership

    We take what worked into everyday use across more teams, sites or systems.

    • Careful rollout with monitoring
    • Access controls and logging in place
    • Runbooks and handover for your team
  4. Ongoing

    Step 4: Make it part of the work

    AI Adoption Partnership

    We train your people, set up governance and keep adding use cases that earn their place.

    • Team training and simple guides
    • Owners, approvals and cost limits agreed
    • New use cases added one at a time
02 · Three ways to start

Find it. Prove it. Adopt it.

Each offer has a fixed scope, a clear end point and a written summary of what you get. Start wherever you are: most teams begin with the audit or go straight to a working proof when the use case is already clear.

  1. 01 · Find2 weeks · Fixed fee

    AI Opportunity Audit

    Find where AI actually pays off.

    Two weeks spent with the people who do the work, followed by a short, ranked list of AI opportunities with honest effort and value estimates.

    • Conversations with operators, managers & IT
    • Review of workflows, data & existing systems
    • Ranked opportunities with effort, value & risk
    Good fit
    For teams that know AI matters but are unsure where to start or which idea to back first.
    You leave with
    A prioritised roadmap you can act on with us, with another partner or on your own.
    Book an AI Opportunity Audit
  2. 03 · AdoptMonthly · Monthly retainer

    AI Adoption Partnership

    Embedded engineers who stay with the work.

    One or two embedded engineers work inside your team: rolling out proven systems, delivering new use cases, training your people and setting up sensible AI governance.

    • 1–2 embedded engineers on a shared backlog
    • Team training & capability transfer
    • AI governance, reviews & documentation
    Good fit
    For teams ready to roll out what worked and keep adding AI use cases, with training and governance alongside.
    You leave with
    Working releases every month and knowledge that stays with your team.
    Discuss an AI Adoption Partnership
Milestone based · Scoped project

Focused System Build

Design and engineering around an agreed scope, from a useful first release through deployment and handover to your team. Scope, milestones and fees are agreed after reviewing your requirements. Hardware work starts with a feasibility check and prototype.

Discuss a focused build

Fees and scope are agreed in writing before any work starts. Any change is discussed and agreed with you first.

03 · The 4-week rhythm

What happens each week, and what it asks of you.

The 4-Week Working Proof runs on a steady weekly rhythm. You always know what we are working on, what we need from your side and what you will have at the end of the week.

Before week 1

A short call to agree the use case, the fixed fee and the success measures. We sign an NDA if you need one. The clock starts once access and a named owner are in place.

  1. Week 1

    Frame the problem

    We sit with the people who do the work, agree the one use case to prove and write down how success will be measured.

    You give

    • A named owner who can make decisions
    • Access to the relevant data, tools and a few users
    • A few hours from the owner and two or three users

    You get

    • A one-page plan: scope, success measures, risks
    • A data and access checklist
    • Confirmation that the use case is worth building
  2. Week 2

    Build on real data

    We connect to your real data and tools and build the first working version. No mock-ups standing in for the real thing.

    You give

    • Quick answers from the owner when questions come up
    • A 30-minute weekly demo slot
    • Sample cases, including the awkward ones

    You get

    • A first working version on your data
    • Weekly demo and a short written update
    • Early read on accuracy, speed and gaps
  3. Week 3

    Harden and check

    We work through edge cases, add the checks a real team needs and test results against the week-1 measures.

    You give

    • Feedback from two or three future users
    • Sign-off on what the pilot will cover
    • The weekly demo slot

    You get

    • A pilot-ready system with access controls
    • Results on real cases against the measures
    • A simple guide for pilot users
  4. Week 4

    Pilot and decide

    Real users try it on real work. We measure what happened and sit down with you to make a go/no-go decision.

    You give

    • A small group of users for the pilot
    • A one-hour decision meeting with the owner
    • Honest feedback from the floor

    You get

    • Pilot results against the week-1 measures
    • A go/no-go recommendation with running costs
    • A rollout plan, or a clear record of why to stop
Next step

Bring one problem. We will tell you honestly if AI fits.

Tell us the workflow and what good looks like. Or browse examples by industry.