Writing

Notes from the operational edge of the AI era.

Featured · Published on LinkedIn

The EU AI Act deadline is 4 months away. Here's the thing nobody is saying: we've already solved this problem.

Risk classification? That's customer risk tiering. Documentation requirements? That's regulatory audit readiness. Human oversight of automated decisions? That's the daily tension between automation and manual review in KYC. The AI governance conversation is dominated by technologists and lawyers. What's missing is the operational layer.

April 2026

Coming soon

The open letter I want to send to my CEO

A meditation on what happens when the platform organization rushes toward AI without the invisible processes that turn activity into profit.

In draft

Coming soon

Five things KYC taught me that apply directly to AI

Risk-based tiering. Policy-to-process translation. Human override by design. Documentation as the product itself. The decade-old playbook the AI industry is about to rediscover.

In draft