About
I’m a senior product designer at Runwise in Chicago. This is how I got here in order, and what each part taught me.
Architecture
2010 – 2016
Separate yourself from the work, and find a faster way to know whether it worked.
I studied architecture at CAPLA, the University of Arizona’s program. The whole thing runs on iteration.
Draw something, pin it to a wall, let a room of people take it apart, then do it again next week. A good number of people who started with me didn’t finish.
What a crit teaches you is that the work is on the wall and you are standing next to it, and those are two different objects. Once that lands, criticism turns into information.
It is also where craft stopped being taste and became a discipline. Every line weight in a drawing set has to be defensible, because somebody is going to build from it.
And a building is a system before it is a picture. You start with the program, which is what the building has to contain, and circulation, which is how people move through it.
I still draw those two first. They go by information architecture and user flows now.
Then I went to work at Booth Hansen in Chicago, on high rises, for exactly one year. Leaving was harder than it sounds. In architecture the profession is the identity. That was five years, a licensure track, and a room full of people who gave up the same things I did to be there.
I liked the buildings. What I didn’t like was the business of it, the pace, and this. You almost never find out whether it worked.
The building opens. Everybody goes home. Nobody comes back to check whether people use it the way you drew.
I wanted the loop measured in weeks.
The side doors
2012 – 2018
Nobody was going to hand me the work, so I built it and showed up with it. That is where the code habit started.
I had no money for a bootcamp. So how do you get in without one?
You build custom WordPress sites for anyone who will hire you, and take marketing jobs next to web apps until one of them lets you do real product work.
I learned to write front end because nobody was going to hand me the design work. And I learned how a business decides what to fund, because I was sitting in marketing while it happened.
Twenty-five or so clients over those years. The ones worth naming: STEM curriculum for a nonprofit that reached somewhere around 300,000 students, the identity and site for a health technology conference that drew 4,000 people to Stanford, and a rebuilt site for Booth Hansen, the firm I had just left, which lifted their client engagement by more than half.
Modest numbers next to a heating system in ten thousand buildings. They are also the first time I watched a design move something I could count, and I have wanted that ever since.
Agencies and in-house
2018 – 2021
Visual craft is an argument about the subject. So is setting the expectation.
The next three years were agency and in-house work. Red Frog, then Adage, then @properties. And in 2020, in the middle of that last one, my wife and I got married. Which that year meant a very small wedding and a lot of hand sanitizer.
At Red Frog the work came with an audience attached. I rebuilt the site for Warrior Dash, a race that puts 100,000 people a year through mud, and it produced their first lift in engagement and ticket sales in five years. I did installations, stage graphics, and signage for Firefly, in front of 80,000. And I ran three design sprints on a ticketing platform to build seated ticketing from nothing.
Type, hierarchy, and whitespace turned into arguments I could defend out loud, tied to what the thing was for. A double border can carry an idea about a building, if you drew it for that reason.
Two things also went wrong here, and both of them stuck. A client wanted revolutionary and we delivered an improvement, which was right for their users and a bad surprise in the room.
And more than once I handed somebody everything I knew instead of the one thing they needed.
Rheaply
2021 – 2023
Most immovable constraints are just badly distributed. Twenty-two squats in a row is brutal and five sets of four is fine.
I spent two years on an asset platform for workplace teams, and most of my time there went to a Google engagement.
It is the most research I have ever done on one project, and it changed the order I work in. Research is how I start now, not a step I get to.
Our daughter was born a few months in. I’ve gotten much better at deciding what needs me since.
I also started writing production code, because the team was small and helping was more useful than waiting.
It added up to 133 contributions in 2022, and eventually to reviewing other engineers’ pull requests rather than only opening my own.
What I got out of that first wasn’t skill. It was finding out what my files were costing the people who had to build them.
And I started weekly design office hours, company-wide. Most of the room wasn’t designers. That’s when I worked out that the part of this I like most is making other people better at it.
Runwise, year one
2023
The screens nobody reviews are where the value is.
I was the first designer they ever hired. The product is wireless building control. Heating, cooling, gas and water monitoring on one device, in about 10,400 buildings.
My users are building managers, supers, and the technicians installing hardware in hundred-year-old basements with no signal.
Being the first designer means you aren’t producing screens. You’re the practice.
So year one was mostly conventions. How work gets reviewed, what a spec looks like, where the design lives, and what quality means here.
I did all of it in the open, numbered v1 through v5, with the questions I hadn’t solved attached. It’s the crit again, with more people in it.
I also designed a set of hardware install flows before I had ever watched an install. They worked. They would have been better if I’d gone first.
Runwise, the middle
2024 – 2025
Belief is the bar, not clarity. And a launch is not a finish.
This is the stretch where I stopped waiting for numbers. I brought analytics to the company and I write my own tracking, so the number that judges a design is one I defined. The gap between a question and an answer became an afternoon instead of a quarter.
Then I spent months solving the wrong problem.
I was sure the problem was whether people could read a number we were putting in front of them. Every figure was correct, and they still didn’t believe it.
Reading a number and believing a number are two different acts, and I had only designed for the first one.
So what does belief take? Not more precision. It took a number from their own paperwork instead of ours.
A launch is not a finish either. I kept watching customers discover features they already had.
And I got steady through churn. My manager described one launch as building the plane while flying it.
What I contributed there wasn’t the artifact. It was staying level, and adding just enough process that we stopped relitigating the same decisions.
Field validation moved to the start of my schedule, instead of something I hoped for at the end.
Runwise, now
2026
I’d rather kill the wrong project than execute it beautifully.
What’s different now is mostly what I argue about. I spent about a year on an exploration, pulled the usage numbers myself, and then made the case to stop. To the head of product, who had been asking me for the designs.
Five years ago I could not have argued my own work off the roadmap.
The rest is AI, and I mean designing it rather than using it. I design and co-build agents that read our repos, answer in plain language, and open pull requests with a person still having to say yes.
The approval gate is the part I’d defend hardest. I don’t think trust in an agent comes from making it more autonomous. It comes from making the takeback obvious.
I’d love to hear what you’re working on. Email me. I answer.