I like taking complex systems apart layer by layer, working out how they actually fit together, and rebuilding them better than I found them. That's what led me into Cloud & Virtualization, where I build clean, automated environments that are designed to handle failure on their own.
Since June 2026 I've been doing this for a living too, part-time alongside my studies, as IT Manager and Tech Stack Maintainer for a group of companies. Most of it is inherited work across stacks I didn't choose, and getting it running again is the easy half. What I actually care about is leaving it structured, documented and automated enough that the next failure doesn't need me.
Three ideas shape how I build: connect the layers, keep the parts independent, and let the system correct itself.
I work across the whole stack, writing automation that reaches over infrastructure boundaries and ties together physical hardware, hypervisors, containers, and routing namespaces.
Systems should be modular, self-contained, and offline-first. Each component needs to isolate, monitor, and recover on its own rather than taking everything else down with it.
Automation shouldn't stop at deployment. I build event-driven and polling watchdogs that audit state continuously and correct configuration drift as it happens, so an environment stays the way it was defined.
I started in the DOS and Pascal era, booting off floppy disks on 286 systems. Learning on the command line shaped the way I still work: break a problem into its structural layers, understand it properly, then decide how to solve it.