skip to content
Developer at a desk split between a retro CRT computer and a futuristic blue cityscape with data visualisations — three hype cycles in one frame.

whoami

37

years of software. Three cycles. One craft.

37 years · 40+ engineers led · 3 hype cycles

Since 2017 I've been Lead Software Architect at Bergx2 GmbH in Munich. I own the architecture of our B2B software there, from long-lived client systems to AI integration that still holds up eighteen months in. Day to day: architecture decisions, coding with agents in the team, and the question of what AI actually makes better for client products and what only fills a slide.

Before that: six years of industrial communication in Erlangen, four years as a tech lead in Switzerland, a short stop at Magix in Berlin, five years in IT security, four years as military police, and earlier still, workstation rollouts for German insurers. Right at the start, in 1989, a first PC that today would belong in a museum.

37 years of software, three hype cycles, one constant: the question of whether something will still hold up in five years. These days I write about that here in the lab and keep building at Bergx2.

→ full track record

how i work

How I work.

01

Quality over speed

Code that's still readable in five years is worth more than code that's done in five days. Speed is a consequence of good decisions, not their replacement.

02

Architecture as foundation

If a system has to hold up, the structural work has to come before the plaster. Every architecture decision goes through the same filter: will this still carry weight in five years, or just produce headlines for one season?

03

Documentation as part of the product

What isn't written down doesn't exist for the team of the day after tomorrow. That goes for architecture decisions, for APIs, for the why behind the how. Documentation is a gift to the next person who opens the code.

04

Visions with a delivery path

A good vision differs from a pretty pitch in one detail: someone can actually build it. Hype is cheap, delivery is expensive. I'd rather plan three steps that land than ten that stay on the slide.

05

AI where it counts

AI is a tool, not a showcase. If it makes the product better, it ships. If it only fills a slide, it stays out. Same for agents in coding: only when they make the team faster, not when they just look the part.

off-code

When the editor is closed.

When I'm not at the code, I'm outside. Hiking in the mountains, exploring new places, travelling without a fixed plan. Nature is the best reset I know, much better than any new framework.

Music is almost always running. While working, in the car, in the evening. And along for everything: my dog Mia, with probably more trail kilometres in her bones than most two-legged hikers.