About.

I am an engineer who writes about software at night, mostly because I do not trust myself to remember anything I have not first written down. Writing is how I think. If I cannot explain something clearly, I probably do not understand it well enough. This site is the long-form version of those notes.

That habit carries over into my work. I have been writing software professionally for about a decade, mostly on the boring backend side of things - databases, distributed systems, tooling, the parts that have to keep working at three in the morning. I like systems that are small enough to fit in one head. The less magic involved, the better.

Coding agents have not really changed that. I use them, and they are incredibly useful, but I still try to build a mental model of every codebase I work on. When something breaks, it is not your LLM's fault. It is yours. The whole agentic coding shift happened so quickly that I am still figuring out the best way to understand software an agent writes. I do not like shipping code I do not understand.

Looking back, I think that instinct started much earlier.

Some of you might remember CheatBook. It was a popular collection of game cheats in the early 2000s. I was less interested in the cheats themselves and more in how a small config change in some .ini file could turn Max Payne into a god, or how a single typo could crash the entire game. Sometimes I would follow the instructions, break something, some changes would be catastrophic and Windows would not boot, and my brother would call the repair guy to reinstall Windows.

A few years later, Ubuntu started shipping installation CDs around the world, and I somehow got one. It felt like discovering a playground. There were endless things to tweak, and if I broke something, I could simply wipe the drive and start over. The repair guy eventually showed me how to reinstall Windows myself, which turned out to be one of the most useful things anyone ever taught me. Around the same time, a local computer magazine included a different Linux distribution with every issue, and I happily tried as many as I could.

By the time I had to choose a college major, the decision had already been made. Computer science was the only thing that made sense.

Not much has changed since then. I still spend most of my time trying to understand systems. The only difference is that these days I type into a TUI, wait for a probability machine to predict the next token, and then try to understand what it just wrote.