About

The short version

I’m Dr. Lukas Pustina — CTO and board member at Scopevisio, Rust developer, occasional speaker, based in the Bonn area. Most of what I’m currently chasing has to do with what coding agents are doing to the craft of software engineering — both for me personally and for the engineering organisation I run.

If you want the longer version, keep reading.

How I got here

My dad introduced me to an Epson PC with an Intel 8088 processor and the GW-BASIC manual, and I produced the kind of spaghetti code that should have been a warning sign for everyone involved. It wasn’t. I kept going. The fascination — that you can describe a thing and the machine will do the thing — has never left.

What’s changed since then is mostly the toolchain. I’ve gone through Pascal, C, Java, Python, Scala, and a brief tour of the LISP family on the way to today, where I write most of my own code in Rust when I get to choose, and in Ansible, Terraform, and the assorted DSLs of running infrastructure when life makes the choice for me.

My first paid IT job was in 1994, managing Linux and Windows NT servers for a local ISP — which is either a flex or a confession depending on your age. I’ve since been the freelancer, the researcher, the consultant, the engineer, the engineering manager, and the executive in roughly that order, and each of those roles taught me something the previous one had hidden from me.

What I do now

CTO and board member at Scopevisio. We build cloud-native ERP, CRM, and DMS software for the German Mittelstand — the kind of customer who has been around for fifty years, whose business is real, and who very much notices when the software gets in the way.

The engineering organisation is north of a hundred people across development, cloud operations, QA, agile, internal IT, IT compliance, and IT security. A non-trivial part of my time is spent on the question of how a company like ours becomes genuinely AI-native rather than just claiming to be on the next slide.

Most days, the job is people. Hiring them, growing them, unblocking them, occasionally disappointing them, and trying to make sure the ones who are here can do the best work of their careers. But I still own the technical strategy, and a strategy you can’t feel in your hands tends to quietly drift away from reality. So I keep my hands in the work.

Keeping a feel for the tech

The thing I’m currently chasing is agentic engineering. Sometime in the last year I rediscovered something I’d half-thought I’d lost: the plain fun of writing software. Coding agents that read, edit, and reason across an entire codebase have pulled me back into the editor at hours I should probably be doing something else. That’s the personal half. The professional half is harder and more interesting: the same shift is reshaping the engineering organisation I run. A lot of my current thinking is going into the downstream questions — what “good” code review looks like when the assistant in the IDE has already read everything, how junior engineers keep growing rather than narrating to a machine, what changes about hiring, architecture, and ownership when the marginal cost of writing code drops by an order of magnitude. I don’t think anyone has clean answers yet.

My testbed is netray.info and the homelab that hosts it. netray is a Rust suite of network-intelligence tools — mhost, Prism, tlsight, ifconfig-rs — written mostly with agents, on a codebase that’s mine and a bar that’s real. None of it is strategically important to the day job. All of it keeps me honest. When I make a call about the production stack at work, I want it to be informed by something more recent than what I built five years ago.

What I believe about software engineering

A few opinions, mostly so you know what you’re in for if you read the rest of this site. The interesting problems sit at the boundaries — between teams, languages, layers of the stack, departments — and most of what I’ve learned that mattered, I learned at one of those seams. Production reality wins arguments; diagrams that don’t survive contact with the on-call rotation aren’t right yet. The craft is changing fast, which means pretending otherwise has a known half-life — and pretending the change is unambiguously good has one too. Old things become new, new things become old; curiosity in both directions tends to outperform conviction in either.

Background

Master’s and PhD in Computer Science from the University of Bonn. Certified data protection officer and information security officer.

Over the years, in various roles: Nokia, T-Mobile, Metro AG, Metro Systems, the German Federal Office for Information Security (BSI), Norman (now part of AVG), Fraunhofer (especially the Cyber Analysis and Defense department at FKIE), and the German Research Foundation (DFG), among others.

About this site

A Gatsby build hosted on Netlify, with a theme based on Alec Lomas’ portfolio. It works, it’s fast enough, and one of these days I’ll get around to migrating it to something simpler. For now, consider the contradiction noted.

There’s a colophon for the curious.