About

I've always liked the messy middle. The product is changing, the brand no longer matches it, the design system is a folder nobody opens, and someone has just asked what AI is supposed to do about any of that.

Those weeks are rarely about a better screen. They're about one system the team can use everywhere — the homepage, the app, the email after someone signs up — so the next page isn't redesigned from scratch.

That's the work I like best.

For 15 years I've worked in that middle: product, brand, the moments a customer actually reads. At Zillow that was Growth Design, then the visual refresh — product, engineering, design, and marketing on one identity.

I still make the things. Leading didn't replace that, and AI didn't either. What's changed is the distance between a decision and a working version. I can put the system in the repo the same week we agree on it. The tool is not the solution. Deciding what to automate is harder than building it, and a team can adopt the tool without changing how they work. A brand is a relationship, not a kit you install. Taste still has to come from a person, especially on the call you can't undo.

My approach

Every project is different. The shape is usually the same.

Frame the opportunity. Who it's for, and what has to be true, before anyone opens Figma.

Build to learn. A prototype beats a debate. The point isn't to be right. It's to find out quickly.

Leave the system better. A screen, a token, a workflow, a habit. If only I can run it, it isn't done.

What I build

Product experiences. Design systems. Brand systems. AI workflows a team can run. Prototypes. A way of working the next person can keep.

Different tools. Same aim: make it clear, build it carefully, and leave the team able to run it.