I design for developers building at the edge of what's automatable. Right now that mostly means figuring out how much control a human should keep once AI starts making deployment decisions on its own. Get that wrong at GitLab's scale and it isn't a bad sprint, it's a customer who stops trusting the platform to ship their code safely.

I started in visual design, then games, then product design. Every version of that job came down to the same thing: take something complicated and hand it back in a shape someone can actually use. I used to do that with posters and board games. Now I do it with pipelines and rollback flows.

Staff Product Designer, GitLab Deploy.

I would rather tell you the flaw in your plan today than let you find it in production next week. I expect the same back.

Finishing something isn't the goal if it's the wrong thing. I've pulled the plug on my own team's work when the research showed it would mislead users, even when the sunk cost said keep going.

I get to the bottom of a workflow before I try to make it bigger. Pipeline behavior, code review patterns, deployment failure modes. I want to understand the mechanism, not just the symptom.

I don't accept "that's not my ethical jurisdiction" as a reason to stop looking at what a workflow enables once it leaves my hands. If something I designed makes it easy to do real harm, that's my problem too, not just a downstream one.

I see the shape of a problem fast, often before the data fully confirms it. Bring me a messy, half-formed problem. That's where I add the most value.

I can take scattered research, conflicting stakeholder opinions, and a vague brief, and find the throughline. Give me the raw inputs, not a pre-digested summary.

I'll tell a PM their pipeline concept will erode trust before it ships, even if it's uncomfortable. Come to me with a specific ask and I'll go deep immediately. A vague one gets vague energy back.

I do my best work in rooms with real tension: competing priorities, unresolved architecture questions, a PM and an engineer who disagree. Put me in that room.

I don't argue with a hard note, I take it. Where I need a beat is turning that acceptance into a visibly different second draft, not relitigating whether you were right.

I'll start ten threads with real conviction and need someone else to close the loop on nine of them. How to help: give me a clear, specific definition of "done" up front. Don't assume I'll intuit it.

I'm better at defining the big swing than at visibly breaking it into shipped pieces along the way. How to help: ask to see the increment, not just the end state, before I've fully worked the whole thing out myself.

I take critique of my work more personally than I'd like to, even when I know it's aimed at the work, not me. How to help: separate the critique from the relationship explicitly if you need to. I recover fast.

Silence from me under stress means I'm regrouping, not that everything's fine. How to help: if I've gone quiet on something that matters, check in directly instead of waiting for me to surface.

The hardest part of coming back after having my son wasn't the caregiving. It was losing the ability to act on my own ideas at the pace I was used to, and needing months to get that back. What actually helped was getting ruthless with my own time, fewer things, chosen on purpose, actually finished.

Growth matters more to me than it would to someone already at the top, and I think about it in scope, not title. What I'm actually after is a shift in which conversations I get pulled into, from "execute this well" to "help us figure out what we should even be doing." If you're deciding who to loop in on a hard strategic question, put me in that room.

Sometimes that means asking in before my current scope has technically earned me a seat. That's not overreach. It's the thing I'm building toward. (I'm working through an Impact MBA on the side toward the same goal, it's a means, not a headline.)

I want options on the table, not one tidy conclusion, because I'll end up stress-testing it myself anyway, sometimes by arguing the losing side just to see if it survives contact. It's part of the job, not me being difficult.

Downstream of me, I need someone tracking the detail I'll lose patience with by day three. And I need people to disagree with me out loud. Saying the uncomfortable thing early is most of what I've built my design career on, and I need that from the people around me too, not just from myself.

Direct, and usually short.

Say it early, say it straight.

100 words means 100 words.

Why do you push back so fast on ideas? Because I've usually already run the idea through a few scenarios in my head before you finished the sentence. It's not dismissal, it's me doing the stress test out loud. If I got it wrong, tell me, I'll recalibrate fast.

Do you actually take feedback well? I don't push back on it, if I'm honest. I accept it pretty quickly. The gap is between accepting something and actually producing something different because of it, and that takes me a beat. Judge me on the second draft, not the first one, that's where the feedback actually shows up.

Do you expect the same directness you give? Yes. I'd rather have an uncomfortable conversation early than a polite one that leaves the actual problem unsaid.

What's the fastest way to lose your trust? Tell me one thing and do another, or hold back information you think I "don't need." I'll extend you the same transparency I expect back.