A five-minute AI maturity self-assessment for leadership teams: where you stand across six disciplines, what is holding you back, and a 90-day plan built from your answers.
Try it live · English and German
Most organizations are not failing at AI because the technology is not ready. They are failing because leaders apply experimentation thinking to orchestration problems. Plenty of pilots, very few at enterprise scale. Tools everywhere, and decisions made exactly the way they were before.
A maturity model is only useful if it leads to a decision. So this diagnostic does two things: it shows where you are, and it tells you what to fix first.
ENGINE treats AI as a leadership and operating model discipline, not a technology rollout. I introduced it in a master track at the GDS Chief Digital Transformation Officer Summit in 2026, where senior leaders assessed themselves against it and named their blockers out loud.
| Discipline | Principle | |
|---|---|---|
| E | Evaluate | Honest diagnosis before any prescription |
| N | Navigate | Governance designed to guide, not to block: responsible speed |
| G | Generate | A new ROI language: Decision Velocity and the Trust Dividend |
| I | Integrate | AI inside workflows, not alongside them |
| N | Normalize | AI-native operations as the default, not a project |
| E | Expand | From enterprise engine to ecosystem standard |
The six disciplines are practiced at the same time, not in sequence. Organizations move through four stages:
| Stage | What it looks like | How ROI is measured |
|---|---|---|
| 1 · Experimenter | Isolated pilots, teams running their own tools | A demo that impressed the board |
| 2 · Enabler | Shared tools and data, AI alongside the work | Hours saved, costs reduced |
| 3 · Orchestrator | AI inside decision workflows, agents with human checkpoints | Decision Velocity |
| 4 · Ecosystem Leader | Standards and trust that partners build on | Trust Dividend and market position |
- The weakest discipline sets the pace. Organizations stall because they are missing two or three disciplines, usually the same two or three. So your stage is your lowest discipline, not your average. The average is shown only as context.
- Name the stall. Each stage has a typical sentence leaders say when they are stuck there. The readout shows it, because recognizing the sentence is often the moment a team admits where it really is.
- A plan built from your answers. The 90-day plan takes your two weakest disciplines and gives one action for each, then ends with a team step: everyone answers separately and compares. Every action is something a leader can start alone. No "set up a working group."
- Content is separate from logic. All framework text lives in
data.js, so the model can evolve without touching the app. - Nothing is stored. No backend, no tracking, no sign-up. Leadership teams are rightly careful about where they type candid answers. Anyone who wants to talk can send their readout from their own email, and they see exactly what they are sending before they send it.
- Bilingual from day one. I work with executives in the US and the DACH region, and a diagnostic should be answered in the language people think in.
Open index.html in a browser. There is no build step and no dependencies.
- Team mode: several leaders answer separately, then see where they disagree
- Printable one-page readout
- An example from practice for each stage
- Tests for the scoring logic
The diagnostic is most useful as a team conversation. When members of a leadership team answer differently, that gap is the finding. I facilitate ENGINE sessions that turn those differences into decisions and a 90-day plan, in English and German. Send me a message on LinkedIn.
ENGINE is a framework I developed for leadership teams moving from AI experimentation to intelligent operations. I am a product leader and former CPO at iHeartMedia. More on my profile, and the operating model behind it in product-operating-system and agent-workflows.