A shared view of the people, delivery phases, operating model, team shape, risks and value measures behind the AI Capability Programme proposal.
1
Stakeholder map
Flowchart
Who backs the programme and who must be won over — grouped by stance, from Champion through Resistant, with the Permanent Secretary as ultimate sponsor and his public April-2027 commitment.
2
Phase 1 approach
Flowchartarc42 · 06 Runtime view
The first-90-days logic: Diagnose → Decide → Prove → Plan & Gate, ending in a go / no-go into Phase 2. Each arrow names the evidence that unlocks the next step.
3
Phase 1 schedule — first 90 days
Ganttarc42 · 06 Runtime view
Mobilise, Diagnose, Govern, Prove and Gate laid over June–October 2026, with the 90-day checkpoint and the use-case decision as milestones.
4
Operating model
Flowchartarc42 · 05 Building block viewArchitecture view
The closest thing to a system architecture: embedded squads, a funded community of practice, and governance that matures with use — a self-reinforcing loop feeding a measurement baseline.
5
Phase 2 sequencing across directorates
Ganttarc42 · 06 Runtime view
Willing-first rollout (Dec 2026 → May 2028): prove with the supportive directorates, scale through the mixed ones, earn trust with the resistant, with a sustainability gate at month 18.
6
Team structure
Org graph
Accountability from the Permanent Secretary down: DBT-side contractual roles, the Zühlke leadership core, the first two embedded squads, and the specialist bench drawn down as needed.
