About

I spent ten years in behavioral health — peer support, crisis systems, federal grants, workforce development. I learned how people hold together under pressure, and how they fall apart. Coherence isn’t abstract when you’re sitting with someone in it.

Now I build mathematical frameworks for AI safety. Not because I switched fields, but because I realized they’re the same field. The patterns that destabilize a person in recovery are the same patterns that destabilize an AI agent under load.

What I’m working on:

• Differential Symbolic Calculus (DSC) — trajectory-based alignment monitoring. Measures whether a system is drifting before it fails.

• World-Model Capital (WMC) — governance protocol for AI agents. Five routing states, cryptographic proofs, structure that makes honesty the path of least resistance.

Background:

• Published in Psychiatric Rehabilitation Journal (APA, 2024)

• $600K+ federal and corporate grants delivered

• Ten years managing cross-system behavioral health programs

What you’ll find here:

Notes on AI behavior, coherence, and what breaks. The peer support lens applied to machines. No mysticism, no hype — just observations from someone watching systems navigate uncertainty.