Field-grade analysis on the work we do.
Regulatory commentary, frontline AML practice, machine learning in regulated environments, and the structural gap in community bank compliance tooling. Written by our founders and engineers, for the people who do this work.
The 34.5% Problem: Why 2024's enforcement surge changes the AML math for community banks
Banking regulator enforcement actions rose 34.5% year-over-year. OCC actions nearly doubled. For community institutions running legacy rule engines, the compliance-cost math has fundamentally changed.
What a BSA examiner actually looks for in a SAR narrative
Twenty years across HSBC, Morgan Stanley, and Capital One told me the same story: the difference between an accepted SAR and a referred one is rarely the transaction — it's the narrative. Here's what examiners actually evaluate.
Why machine learning in AML is not optional anymore — and why 'explainability' stops being a reason to delay
The reflexive "regulators won't accept black-box models" objection was a reasonable concern in 2018. In 2026, it misunderstands both what examiners accept and what explainability means for case-level decisions.
The community bank AML gap: Why the institutions most exposed to financial crime have the fewest tools to fight it
Three structural reasons the less than $50B asset tier has been systematically underserved by AML technology — and why the economics finally work in 2026.
What banks see alone vs. what banks see together: The case for secure collaboration in AML
A $40K structuring operation spread across eight banks at $4,800 each. Every institution sees a slightly unusual pattern. None, alone, can detect the coordinated activity. The full picture only exists in aggregate.
Questions on anything you've read?
We answer every message directly. Whether you're a practitioner, an investor, or a fellow builder — reach out.