Complex products for expert users. Users in the iteration loop.

The work

Hard domains, expert users, real stakes when you get it wrong.

I work on products where the domain is genuinely difficult and the interface has to disappear without losing the people who depend on it. That spans trustworthy AI, enterprise workflows for users who already know what they need, and public-scale surfaces with no margin for confusion.

The job is figuring out what should be built before build spend locks in the wrong direction: direct access to users, evidence over artifacts, and the willingness to recommend against an idea when the signal says no. Proof and depth live under Work, organized by domain, practice, and outcome.

How I work

Honest user feedback closes the loop.

What I optimize for is the compression between a design call and feedback from people who depend on the product. Functional code is how I put ideas in front of them early, before decisions harden. Prototypes that increasingly ship, whether the domain is clinical voice AI, a security investigation platform, or a solo system where uncertainty is the product condition.

Full-process product design goes deeper on discovery through delivery, divergent research, and when to say no.