About

I design the system before I choose the stack. That is the part that transfers.

I started building things at ten, which meant I picked up the habit of learning a technology because a problem needed it, rather than because it was next on a syllabus. That is still how I work.

The result is a range that looks scattered on a CV: firmware and Flutter, scrapers and Shopify, games and reverse engineering. What connects them is not the tools. It is that each one was a system with a source of truth, a failure mode, and a data path that had to be designed before anything could be written.

The fair version of the worry is this: someone who has touched ten things may know none of them deeply, and may leave when the job turns out to be one thing.

My answer is that the depth is on a different axis. An offline sync conflict, a dropped MQTT uplink, and a rate-limited scraper are the same question in different clothes — what is authoritative, and what happens when two sources disagree. I have answered that question in enough shapes that the framework is the easy part.

I take unfamiliar projects deliberately, because that is the fastest way to find out where my model of a problem is wrong.