CodeCrafts 2025

Conference Report #

I went to CodeCrafts 2025 in Vienna to learn, to meet people, and to get new energy for building software. This year was the 4th iteration of CodeCrafts, again hosted by SQUER—and at this point it really feels like a stable event in the Vienna tech landscape.

Just like last year, it took place in the bigger Expedithalle, and overall the day felt very similar to CodeCrafts 2024, with the same ups and downs.

Venue & Vibe #

EXPEDITHALLE is large, and the bigger stage worked well: good visibility, good sound, and a smooth “real conference” flow. At the same time, with more attendees, conversations became a bit more short and fast.

If you come mainly for deep hallway-track discussions, you can still get them—but you have to work harder: find quiet corners, pull people aside, or commit to longer chats rather than quick “nice to meet you” exchanges between sessions.

What I still really liked: despite the size, the conference felt community-driven. People were open, and the talks didn’t feel like marketing pitches. Many sessions connected architecture decisions with organizational reality, which is exactly where most “clean diagrams” get challenged.

Theme consistency #

The overall theme stayed consistent with the last years: DDD-ish thinking, architecture, quality, and system evolution, plus the constant tension between ideal design and what organizations can actually sustain.

Speakers were again high quality, but compared to 2024, it was slightly less interesting to me personally—not bad at all, just fewer “I need to change what I do on Monday” moments than last year.

Food & after-party #

Food was good (it matters), and the after-party was good as well—a relaxed ending that made it easier to have more honest conversations than the usual “conference small talk”.

Main ideas I took home #

1) Software is about choices, not only tools #

Kent Beck opened with “The Forest & The Desert Are Parallel Universes”. For me, the message was that different contexts need different approaches. Some teams work in a “forest” with many paths and experiments; others are in a “desert” where survival and simple rules matter most. I liked this because it reduces pointless debates. Instead of saying “this method is always best”, we can ask: what kind of environment are we in right now?

2) AI and architecture: helpful, but not magic #

Avraham Poupko’s talk about using AI to create software architecture was interesting because it did not sound like hype. My main takeaway: AI can speed up exploration and documentation, but you still need strong thinking about trade-offs, constraints, and long-term ownership. If you don’t understand your system, AI output can also make confusion faster.

3) Security basics are still not “basic” #

The web app security talk by jackie (Andrea Ida Malkah Klaura) was a good reminder that many problems are not new, but they still happen. I left with the feeling that security is less about “secret tricks” and more about everyday habits: validating inputs, managing auth correctly, and treating security as part of normal development—not a late checklist item.

4) Legacy systems can evolve with the right flow #

Susanne Kaiser talked about evolving legacy systems with “architecture for flow”. I liked the focus on changing the system in a way that supports how teams work, not only how code is structured. The idea is: reduce friction, shorten feedback loops, and make safe changes easier. This is a very realistic view of legacy work—because most companies cannot “rewrite everything”.

5) Quality vs. speed is a false fight #

Adam Tornhill’s talk on code quality and empirical insights hit a nerve. The “we need to go fast, so we can’t clean up” argument is still common. The talk made a strong case that poor quality often slows you down later, and that you can measure parts of this instead of guessing. I also liked that it connected engineering decisions to real outcomes, not just style preferences.

6) Event sourcing: practical steps matter #

Shahab Ganji gave a practical view on event sourcing. I appreciated that it was not presented as the answer for everything. The talk helped clarify where event sourcing fits well (auditability, complex domain history) and what you must be ready for (new complexity, event design, and operational concerns). I left thinking: if we do it, we should do it deliberately and with training, not as a quick experiment in production.

7) Sustainability can be a refactoring topic #

Bjorna Kalaja presented refactoring techniques to reduce the environmental impact of software. This was one of the most refreshing talks for me. It made sustainability feel like an engineering topic: optimize heavy operations, reduce waste, and design systems to use fewer resources. I liked that it was concrete: you can start small and still make a difference.

8) Architecture as Code: make decisions executable #

The “Architecture as Code” session with Neal Ford and Mark Richards connected architecture work to automation and developer workflows. My main point from this: architecture decisions should not live only in slide decks. If we can encode rules, checks, and standards into pipelines and tooling, we reduce “architecture drift” and avoid relying only on human memory.

Networking & Atmosphere #

With 500+ attendees, the conference was big, but the single track and shared breaks still made it feel like a shared experience. At the same time, the size pushed networking a bit toward breadth over depth: lots of quick check-ins, fewer long conversations—unless you intentionally made space for them.

Final Thoughts #

CodeCrafts 2025 delivered what I wanted: practical ideas, mindset shifts, and a feeling that many people still care about building software well. It also confirmed something I already felt in 2024: CodeCrafts has a clear identity (architecture/DDD/system thinking, grounded in reality), and it’s becoming a fixed point in Vienna’s conference calendar—especially thanks to SQUER keeping it community-first.

If I compare it to last year: same spirit, same strengths, same trade-offs. Great venue and execution, slightly harder to get deep “hallway track” time, strong speakers overall, but for me personally a tiny bit less “wow” than 2024. Still: I left Vienna with notes, new contacts, and a clearer plan for what I want to improve in my own work.