Every institution has one — currents of capital moving constantly, most of it clean, some of it carrying what shouldn't be there. My work at Morgan Stanley, leading Global Financial Crimes efforts in AML, sanctions and compliance, has been about learning to read that current: tracing its source through enhanced due diligence, recognizing what doesn't belong, and building the frameworks that keep the whole system trustworthy.
For years, that meant refining how well I could read the water myself — earning a CAMS certification, closing control gaps, and being recognized for the rigor that work demands.
But the more interesting shift has been moving from reading the river to reshaping the banks. Alongside my team, I helped build a Policy & Procedure Agent that gives our Client Risk Unit a constant, reliable reference point — no one navigating from memory alone. We also built an ACAMS Study Agent to help our team prepare for certification, and it's delivered a 100% first-attempt pass rate for two consecutive years.
Neither project was about replacing judgment. Both were about giving good judgment better tools — freeing people to focus on the parts of this work that are irreducibly human: discernment, mentorship, and the instinct to ask "does this belong here?"
That's what excites me most about where agentic AI is headed in compliance. The current isn't slowing down. The question is whether our systems for watching it can keep pace — and I believe they can. Recently I earned the Agentic AI 2.5 Week Intensive Certification from Harvard Data Science Review, and I'm pursuing the Data Science Track Career Certification with DataCamp.
