Leading AI adoption across a design organisation
Meeting people where they are: a three-tier path that took a design team from curious to daily practice.
The problem with "everyone use AI"
Most AI rollouts fail in one of two ways: a mandate nobody follows, or a handful of enthusiasts racing ahead while everyone else watches. Neither changes how a team actually works. I wanted the capability to belong to the whole design organisation, not to me.
A path, not a mandate
I built a three-tier approach that met people where they were. An accessible entry point for the wider team, a familiar middle tier inside the tools designers already used, and a power tier as the north star for those ready to build with code. Each tier had training, examples from our own work and a clear next step.
- designers using AI in their daily workflow
- 30+
- tiers, from first steps to building in code
- 3
Governance that enables
I led the design and creative side of the company's Claude rollout: training, tooling and the guardrails that kept output on-brand and accessible. Because our design system rules were codified as data rather than wiki pages, AI tools could follow them, which made governance something that ran automatically rather than something people policed.
I also represent design on the company AI Council, so design moved from reacting to AI decisions to shaping them.
Teaching teams to run it themselves
The test of adoption is whether it spreads without you. I ran enablement sessions for other teams, walking them through our guardrails and how to build and prototype the same way, then handed over the examples and set-up so they could carry on alone.
What changed
The design team now works with AI as a matter of course, and external benchmarking confirmed we were ahead of the curve on AI maturity. More importantly, the practice no longer depends on any one person.