Small AI tools, public-safe notes, and prototypes that turn messy context into inspectable decisions.
About
I work on the ordinary layer where AI meets real work: what gets remembered, what gets checked, what should stay private, and how a messy transcript or note becomes something a person can inspect.
About facts
How memory, evaluation, voice notes, and human review can help judgment without becoming identity capture.
Private conversations, employer context, family or health details, financial account material, or identity capture.
Public boundary
This site can show methods, synthetic examples, public repositories, and the shape of my thinking. It does not publish private conversations, employer material, family or health details, financial account material, or unapproved identity traces.