AI adoption strategy for a 5,000-person workforce
Building a portfolio-level roadmap, governance model, and enablement program that moved AI from scattered experimentation to operational use.
The situation
- Over 40 disconnected AI experiments were live across business units, with no shared view of value, risk, or duplication.
- Leadership wanted a credible 18-month plan that the board, partners, and regulators could all sign off on.
- Internal capability was thin — a handful of strong individuals, no operating model to scale them.
How we engaged
Discover
Interviewed 35 leaders, catalogued every active experiment, and quantified the addressable value pools across service lines.
Design
Produced a tiered portfolio (foundational productivity, augmented delivery, new offerings) with sequencing, investment cases, and a governance operating model.
Deliver
Stood up a Centre of Enablement, codified usage policies, and launched the first three lighthouse implementations with embedded delivery.
Adopt
Designed a tiered enablement curriculum, partner-level office hours, and a value-realisation cadence reported to the executive committee quarterly.
What we built and ran
- An AI portfolio map across three horizons with explicit owners, KPIs, and investment cases
- A governance model covering acceptable use, model risk, data handling, and vendor evaluation
- A Centre of Enablement charter with capability roles, intake process, and shared platform standards
- An enterprise-wide enablement curriculum tiered by role and risk exposure
What we'd carry forward
- Without a portfolio view, AI investment fragments into vanity pilots. The roadmap is the unlock.
- Governance is enablement when it's well-designed — it tells people what they can do, not just what they can't.
- Lighthouse implementations matter more than slideware. Ship three real things in the first 90 days.
Want a tailored example for your context?
Tell us about your industry and the outcomes you're after. We'll send a relevant scenario and a recommended path forward.