One discipline, held from tactical collection to the C‑suite.
I lead AI governance, security, and risk for organizations where being wrong is expensive — in technology, healthcare, and the regulated and high-consequence sectors next to them — so powerful systems get used properly, at speed.
My work has been security, risk, and governance for fifteen-plus years, without a break — and for most of them, director and senior leadership building the data and AI capabilities that work protects: the translation layer between heavy engineering reality and executive intent.
As AI systems move from static models to autonomous agents — writing to systems of record, modifying access controls, integrating with clinical and industrial hardware — the central challenge stops being capability. It becomes governance at scale, and the data architecture beneath it. Without sound platform foundations, governance is policy without plumbing. Without enforceable governance, platforms generate risk at the speed of automation.
I design the architecture that solves both: the decision rights, autonomy limits, and evidence records that let an organization delegate consequential action without losing control of it. This is the work organizations increasingly need at the top of the house — the VP of AI mandate, the Chief AI Officer charter, the standing governance board. Solve liability and compliance at the control plane, build the data infrastructure that feeds it, and rigorous governance becomes a velocity enabler rather than a brake.
That claim is live, not theoretical: alongside the advisory work I currently run an enterprise AI-literacy program — standing up the bench of human judgment that proper use of these systems depends on. Systems don’t hold judgment; benches do. Enablement practiced, not preached.
The spine
I never left security. I have operated where system failure has consequences well beyond the enterprise — and I still do: the current substance of the work is the security and governance of AI itself. It informs everything I design, and I build accordingly.
Before I governed AI, I shipped it.
The governance thinking is earned, not theorized. For two decades the assignment has had the same shape — walk into an environment with no capability and stand one up, at increasing scale and altitude:
- A large-scale NLP and data-science capability, built from zero inside a nine-figure program.
- A data and AI practice scaled roughly eightfold — from low single-digit millions to $25M — with margins nearly doubled, from ~30% to ~50%.
- Delivery across multi-hundred-person organizations and agency-wide deployment scope.
I design oversight the way I do because I’ve carried the P&L on the other side of it. Governance written by people who’ve never shipped is the theater I’m arguing against.
Why defense is the proof, not the pitch
Almost no one competing for these seats holds the whole record — and the record is warrant, not backstory. Every claim on this site has a receipt: collection run under operational consequence, capability built from zero, a P&L carried, an enablement program running live today. The governance problem is identical in shape whether it sits in a hospital system, a bank’s model-risk function, or a federal program: the comp ceiling differs; the discipline does not.