Work

Five case studies

Each one opens with the leadership question it answers, because decisions, not deliverables, are what a leadership role runs on. Two studies are open; the rest are shared on request: email me for the password.

• 01AI research you can interrogate How do you scale a research practice when demand doubles, without hollowing out the thing that makes research worth having? Synthetic personas grounded in real participants: provenance to source transcripts, a confess-ignorance contract, and an eval suite guarding the grounding.
• 02Prototyping the strategy before it had a nameon request What do you do when you can see where the company needs to go, and it isn't your lane? 50+ open-source MCP servers for pharmaceutical data, built months before agent-accessible intelligence became company strategy. And the choice that mattered: fighting for the thing, not the ownership.
• 03Helix: winning the design system fight How do you build contested infrastructure, with no mandate but the one you create? From nothing to the portfolio standard in two years: more than ten products on Helix, legacy tools migrating and new ones adopting it by default, requested unprompted.
• 04Research as an operating systemon request What does it take to build a discipline from nothing, and then make it the way the organization decides? Founded the practice, hired every researcher, moved it upstream. Demand up 114%, four business units, and the alignment peak landing in the quarter of deepest expansion.
• 05The AI validation gateon request When the whole company is mandated to ship AI, how does design leadership keep it honest? Research repositioned as the answer to AI uncertainty: 24 validation workshops, nine AI products deeply validated, and checkbox AI stopped before it shipped.