The program

A delivery program, with training inside.

Every AI Champion delivers one real business workflow in 12 weeks and enables their team around it. Useful, tested on real examples, measured against a baseline, approved against simple governance rules.

The format

Built to be light on your people and heavy on outcomes.

12 weeks

One cohort, one clear arc from intake to a management demo.

Around 10 Champions

Small cohorts, selected from business teams, not developers or AI engineers.

8 hours a week

Advised commitment from each Champion, alongside their day job.

High-impact workflows

Low complexity, repeated weekly, with a clear owner. Complex use cases go to the roadmap.

Model agnostic

Claude, Copilot, ChatGPT, or Gemini. The value sits in the workflow.

Management in 3 moments

Kickoff, midpoint check, and a final demo of real results.

Week by week

Twelve weeks, each with one output.

No abstract theory. Every week moves the workflow one concrete step closer to shipped.

01
Intake and baseline
Current tools, pain points, and a shortlist of workflow candidates.
02
AI foundations for business users
Safe usage, tool options, and what makes a good or bad use case.
03
Use case selection
Each Champion picks their high-impact, low-complexity workflows. Complex ones go to the roadmap.
04
Workflow mapping
Current process, bottlenecks, data, and risk.
05
Prompt and workflow design
A first working version, built on real inputs.
06
Tool and connector setup
Controlled use of the chosen AI client, with approved access.
07
Testing with real examples
Quality, time saved, and failure modes, on actual work.
08
Governance review
Data rules, approval, and usage guardrails.
09
Pilot with a small group
The workflow used by colleagues, not just its author.
10
ROI measurement
Time saved, cycle time, quality, and user feedback.
11
Final demo preparation
The story, the numbers, and the next steps.
12
Management demo and scale roadmap
Next use cases, cost forecast, and ownership.
Champion deliverables

What each Champion walks away with.

Personal AI workflow

One practical workflow running in their own department.

Before and after maps

The current process and the AI-assisted version, side by side.

Instruction pack

Reusable instructions, prompts, examples, and quality criteria.

Tool choice rationale

Why this AI client fits this workflow.

Data and risk checklist

What data is used, what is avoided, what needs review.

Pilot evidence and ROI

Test results, feedback, and a payback estimate in EUR terms.

Program rules

Seven rules keep it honest.

The rules are what stop the program drifting into a masterclass or a pile of demos. They are non-negotiable.

How we choose what to build

We start where AI pays back fastest.

Every use case is scored on business impact and complexity. Champions build the quick wins now. The big, complex bets are logged for the roadmap, to pursue later through outsourcing, procurement, or an RFP.

Complexity

Avoid

Lots of work, little return.

On the roadmap

Strategic projects

Big impact, long haul. For later.

Low priority

Nice to have, not now.

Quick wins

High impact, low complexity. The ideal use cases.

Where we start
Business impact
Choosing the workflow

What makes a good workflow.

✓ Good

  • ✓ Repeated weekly
  • ✓ Clear owner
  • ✓ Existing pain, not nice to have
  • ✓ Low to medium risk
  • ✓ Data already accessible
  • ✓ Measurable against a baseline

✕ Bad

  • ✕ One-off task
  • ✕ No owner
  • ✕ High risk from day one
  • ✕ Needs complex integration first
  • ✕ Vague productivity claim
  • ✕ Custom build dressed up as a workflow
Next cohort opens Q4 2026

Pick the workflow. We deliver the capability.

Book an AI workflow scan See the outcomes
Small cohorts · high-impact workflows