Approach
How I think about AI in the enterprise.
Not a service menu. A territory — the questions I keep returning to when AI ambition meets the reality of a regulated, complex organization.
Most enterprise AI doesn't fail on the model. It fails on the operating model. The work is building the system underneath: an AI governance framework that enables speed rather than blocking it, coordination across the functions that have to deliver together, and the living knowledge layer the whole thing runs on. That's the difference between an enterprise AI strategy that looks good in a deck and one that actually changes what the work produces.
- 01
Governance as enabler
AI governance designed to accelerate good decisions — not to slow every decision down. Guardrails that scale with ambition, not against it.
- 02
Alignment across functions
Strategy, technology, legal, medical, commercial — all using the same words, the same definitions of value, and the same view of risk.
- 03
Decision-making and judgment
Where AI assists, where it decides, where it stays out. Designing the human-in-the-loop so judgment compounds instead of erodes.
- 04
Knowledge systems underneath
Content, data, and context engineered so AI can actually use them. The unglamorous foundation that determines whether anything else works.