Find Your Footing

I've been building software for the last 15 years. I've worked on big teams and for small companies, I've launched projects from my home office, and I've failed at building far more things than I've shipped. I've not enjoyed every minute of it, but I have loved the journey.

A year and a half ago, after nearly seven years, I left my gig as Design Manager at Elasticsearch. Health issues caused me to take a step back from the grind to re-evaluate what I wanted to spend my time doing. I started kicking up dust on a project that I'd been mulling around in my head for a number of months. I wanted to build a school management and billing tool for my kids' private school. They were living in the dark ages, and the idea sounded challenging in all the right ways and like I'd be providing something that mattered.

In March of '25, I wrote my first lines of code for Lazarus. I spent the next several weeks as I had so many nights and years before: scouring Stack Overflow for answers to problems I'd never encountered. I dug in and followed my same routine of designing in the browser, building requirements as I went, but being cautious about what I spent time writing production code for, because every line cost me real energy and time.

After about two months, something changed. Claude Code became generally available. I installed it, pre-paid for some tokens, and slowly found myself replacing those Stack Overflow visits with questions in the terminal. I found that I instead of 100%, I was now only writing about 95% of the code that I needed. I was still penning the GraphQL mutations by hand. Every line, every missing end, every empty dependency array in my useEffect(). I felt like I was soaring, though. There was no way that something could be so good at being my coding assistant. It felt like I'd travelled to the future.

My first experience with something like this was the GitHub Copilot preview. I was working for Elastic on our open-source design system. Because the code was open source, Copilot could rip out a pretty well-written data table layout in a few seconds. It was fascinating, but it was obvious.

This, though... this felt magical. We weren't writing "guessed" React components; this was me describing a real issue and getting a tailored solution. It was fascinating, and it felt mysterious.

We all know what the next 18 months have looked like. The craziest part to me is how fast it happened. In the last three months, I've not written a single line of code in Lazarus or the three other projects I actively contribute to.

A few nights ago, I built a dispatcher system that runs on a desktop at home and listens for Basecamp issues assigned to a persona. When one comes in, a headless Claude session starts and investigates the issue. Follow-up questions are automatically posted in Basecamp with the author tagged. Fixes are automatically built, and PRs are opened for review. The whole system tracks its approximate token spend per task and shares real-time updates to my other computers over Tailscale and a custom Omarchy plugin.

By the end of the first night, Gary (the dispatcher) had opened 23 PRs for review. A day later, after code reviews, testing, and only two changes, my partner and I merged them all and shipped.

I built the entire system in 3 hours while sitting on the couch watching I, Robot with my son.

I'm not a great engineer, but I'm a really good problem solver. And I think that's the era we're in right now. For years I've scratched and clawed at projects trying to get things built in my basement. Things would stall, I'd lose interest, and inevitably something else would catch my attention. Another problem to be solved.

This is an entirely new paradigm. I can build almost at the speed of thought. And my background as a designer, engineer, and product manager has allowed me to finally wear all of my hats well at the same time. There will be folks who struggle to adapt to the changing tide. But there will be others, maybe like myself, who have struggled to find their real calling in this industry and now find themselves empowered to create like never before. That's a really freakin' cool thing.

So where's it all go from here? I don't know. I imagine there's gonna be a lot more solutions to a lot of different flavors of the same problems. Code has never been cheaper. It's never been easier to go from idea to "usable tool." People are going to build the things they need. The audiences will be smaller; the solutions more bespoke. It's gonna get weird. And it's gonna be so fun.

I'm currently working with a few very small teams, and we're building some very cool things very quickly. And the most incredible part is that none of it is crap. It's all highly useful, requested stuff that our clients are asking for. Things that have sat on the backlog for years because there was never enough time to get to them. Large feature requests that used to compete for priority are now being built in parallel with the things we believe are gonna help us grow.

Folks who don't find a way to embrace the change are going to be left behind. Folks who try too hard to hold onto their titles and their responsibilities of yore will miss out on getting to see how their skills, applied differently, can open up entire dimensions they'd never seen before. The ground has shifted. This is your reminder to find your footing.