Twelve questions about your organization across five dimensions of AI maturity: Strategy & Governance, Platform & Engineering, Data & Knowledge, People & Operating Model, and Use Cases & Value. The radar fills in as you answer. When you're done, one structured-output call generates a sequenced three-phase roadmap with concrete actions tagged by role and effort — the shape of a real consulting deliverable, in 30 seconds.
Each question contributes to one dimension on a 0–4 scale. Dimension scores are simple averages of contributing questions you've answered. The 0–4 scale loosely tracks: 0 nothing, 1 ad-hoc, 2 defined, 3 repeatable, 4 optimizing.
When you click Generate roadmap:
1. Compute dimension averages from answered questions
2. Snapshot: { industry, org_size, ambition, scores, answers }
3. POST /api/lab/chat with a structured-output prompt
- system: "You are an AI advisory partner. Given this snapshot,
return a sequenced 3-phase roadmap as JSON…" (versioned: maturity.v1)
- user: the snapshot
- response_format: json_object
4. Parse + render — executive summary, strengths, gaps, three phase cards
with concrete actions tagged by role and effort.
Single LLM call. The radar chart and dimension scores are computed client-side as soon as you start answering — no model needed for those. The model only writes the prose recommendation, given structured numbers as input.
Generate a roadmap to see telemetry.
This is a 12-question demo, not a 200-page consulting engagement. A real maturity assessment includes interviews with named owners across business and IT, document review, sample artefact inspection, and a benchmarking layer comparing the org to industry peers.
What this demo gets right: the shape of the deliverable. The five dimensions are the real ones, the 0–4 scale is the conventional one, and the three-phase roadmap is the structure most CIOs expect. Use the output here to start a conversation, not to commit to a plan.