Skip to main content
If you scroll through my projects, they don’t obviously belong together. A Linux kernel patch. A humanitarian mapping tool. A car scraper. An escape room solved by evolution. A batch script that sorts your Downloads folder. Different languages, different scales, different stakes. But there is a thread, and it’s this: I build to understand.

​
Curiosity first

I really like learning about different topics and finding communities to grow with — that’s not a tagline, it’s genuinely how I pick what to work on. A genetic algorithm sounded abstract, so I made one escape a room until I could see it learn. I wanted a used car, so I turned the search into a dataset. My mouse didn’t work on Linux, so I read enough of the HID subsystem to fix it for everyone with the same mouse. Each project is a question I couldn’t answer by reading about it. Building is how I close that gap.

​
Small is fine

The most-starred thing I’ve made is a ten-line script. The thing I’m proudest of is a six-line kernel patch. I’ve stopped measuring a project by how impressive it sounds and started measuring it by whether it taught me something or helped someone. Usually the small ones win on both.

​
Out in the open

Almost everything I do is public, and a lot of it is contributing to work that isn’t mine — OpenStreetMap, the kernel, HOT. Building alone teaches you the tool. Building with a community teaches you the craft. I’ll take the second one every time.