Over twenty years in mortgage risk. A new chapter in AI-assisted work.

For over 20 years, I worked across nearly every side of the mortgage industry — sales, underwriting, quality control, and risk. That range gave me a close view of how small inconsistencies in judgment or process compound into real financial and compliance exposure at scale.

That background is exactly why the current wave of AI-assisted evaluation work feels familiar. I'm currently building a Credit Risk Playbook — a company-wide framework that uses AI to help identify and address risk-related issues before they become costly — bringing structured, underwriting-grade thinking to a problem space that's moving faster than most risk teams can keep up with.

That work happens hands-on, not from the sidelines. I build and test directly with Claude, Gemini, and Copilot across these projects — which means the evaluation instincts I bring to RLHF and data-quality work come from actually using these tools daily, not just reading about them.

Apex Prism Concepts is where that same discipline shows up in smaller form — careful, high-precision data evaluation and RLHF work, done with the same attention to detail and compliance mindset built over two decades in a heavily regulated industry.