About me

I work across the whole stack, from the database and the API to the interface people actually use. Before I write code I like to understand the problem and sketch the solution, and I leave tests behind so the next change is safe.

AI coding agents make me much faster in my daily work. I plan the work, build it with their help and review every change, the same way I would review a colleague's code. I work with Claude Code and Codex, using agents, skills and MCP servers.

The details users feel without noticing them: a page that loads faster than expected, search that understands a typo, a transition that shows where you came from, a form that works with the keyboard, an app that keeps working offline.

I also like the work behind it: finding why a page is slow, putting an app into a container and onto a server, and making accessibility and SEO part of the build instead of an afterthought.

I studied computer science at the Faculty of Informatics and Information Technologies in Bratislava and then intelligent software systems, where my master's thesis used neural networks to help detect glaucoma in eye images.

My first job was in Python, building data pipelines for a large e-shop, where I also learned Ruby for an automation task. Then I built a REST API and React app as a full-stack developer, and at IGT Systems I learned Elixir and Kotlin Multiplatform for new projects. Picking up a new stack when a project needs it has become a normal part of my work.

I run my project demos on my own home server with Docker, Tailscale and monitoring, and their data resets every night. Every December I solve Advent of Code, each year in a different language. I do sports to balance the screen time, and I am relearning German.