The publication
Essays on operational resilience, scenario analysis, and the widening gap between regulatory expectation and bank tooling. Written from inside the discipline, not about it.
Read the essays
An independent publication & prototyping lab
Operational risk is being rebuilt around resilience and AI. The rebuild is documented and prototyped here.
I spent 18 years inside North American banking risk functions: stress testing, CCAR, operational resilience, regulatory engagement. oprisk.ai is where I publish the research and build the tools the discipline should already have.
The Lab
Working prototypes, not products, that blueprint what modern operational risk tooling should look like for mid-tier banks. Two flagship modules are live: the first components of AEGIS, a unified eGRC platform of twelve modules built to work as one.
Explore the LabThe OpRisk Signal
Signal, not noise: a practitioner's curation of what actually matters in operational risk and AI, with one essay in every issue. Written for risk leaders, read in one sitting.
Subscribe to the Signal
Featured build
ORBIT
A working prototype for operational resilience: important business services, impact tolerances, dependency mapping, scenario testing, and board-ready reporting in one workbench.
Start here
New to oprisk.ai? The founding essay lays out the thesis and the full research agenda: why operational risk tooling stalled, and what closing the gap looks like.
Read “Why I’m building in public”
Latest
Recent writing
The build-versus-buy inversion in risk technology
For twenty years the only honest branch was buy. The premise that made it true has expired, and the widest build-versus-buy gap in risk technology is still widening.
Read the essaySplicing the tail: the severity modeling choices that actually matter
The body of the severity distribution is easy; the tail is the entire game. A walk through the modeling choices that actually move the answer: splicing, thresholds, and the copula default nobody questions.
Read the essayThe second line doesn't need fewer people. It needs instruments.
Every AI conversation in risk starts with the wrong question: headcount. The second line's binding constraint has never been intelligence. It's attention, and attention is exactly what the new instruments return.
Read the essay