Tech and Meet: Event Sourcing And CQRS For Dummies

27/05/2025

On May 27, 2025, I went to a Tech&Meet session at Howest Brugge, the session was called "Event Sourcing and CQRS for Dummies," and it was led by Kim Van Renterghem. Kim's a senior dev over at zf-scalar, and he broke things down in a really digestible way. He talked about how standard CRUD setups can't really preserve the history of changes, which ends up being a pain when you need to debug or audit stuff later on.

Instead, he walked us through Event Sourcing. He treats every change in the system as a distinct, unchangeable event, like "DrinkOrdered" or "TableReserved." These events pile up over time and basically tell the whole story of what happened, so you can always recreate the current state from scratch if you need to.

To kick things off with Event Sourcing, Kim recommended using Event Storming. It's more of a collaborative modeling thing where the team gets together to sketch out the domain, mapping events, commands, people, and groupings. Super helpful for making sure everyone's on the same page, even outside the dev team.

He also touched on CQRS, which stands for Command Query Responsibility Segregation. It separates how you write data from how you read it. So whenever something happens, it updates the write side and then projectors sync the read side. This setup makes scaling smoother and lets you build solid analytical dashboards without clogging up the operational side.

Sure, these patterns add some extra complexity, and they take discipline to pull off. But the benefits like full history tracking and being able to replay past events, make them worth it in systems that need to grow and adapt and can possibly use this technique.

Podcast Episode: From Mechanics to Data: The Journey of a Data Engineer

On May 23 I got to co-host an episode of The Howest IT Student Podcast with Jellis and Jensen. Our guest Ben Samyn is a data engineer with a pretty unique background. He actually started out in electromechanics. The episode "From Mechanics to Data: The Journey of a Data Engineer" dove into his path and touched on a bunch of bigger-picture topics like AI, data privacy and how tech is changing things in the energy world.

Prepping for the podcast was a learning experience in itself. We had to make sure the chat stayed natural but still hit all the main points we wanted to cover. We sometimes had to ask follow-up questions on the fly and figure out how to nudge the convo back on course without it feeling forced. It really gave me a feel for what makes a good interview flow.

What stuck with me most was Ben's story. Switching from mechanical work to data engineering isn't exactly a small pivot but his journey showed just how valuable transferable skills can be. It's a solid reminder that career paths aren't always linear and that it's totally possible to start fresh if you're willing to learn.

I also really liked our chat about how data is used in the energy sector. Ben broke down complex stuff like balancing the energy grid or using automation tools in a way that made total sense even without a deep technical background.

And then there was our discussion on AI. Ben had a grounded take how "AI" gets thrown around as a buzzword way too often and how companies are slowly learning what it actually means. That part really made me think about how hype can sometimes get in the way of doing real impactful work.

All in all co-hosting this episode was something I'll remember. It was fun insightful and gave me a ton to take with me going forward. If you speak Dutch I'd definitely recommend giving it a listen.

Tech and Meet: Infrastructure As Code

20/05/2025

On May 20, FlowFactor hosted a hands-on session all about Infrastructure as Code (IaC), with a strong focus on optimizing Terraform workflows. they really leaned into the value of staying disciplined when managing cloud infrastructure.

One of the key things they stressed was how important it is to manage Terraform's state file the right way. Since this file maps your code to the actual resources, it should really be stored remotely using backends like AWS S3 or Terraform Cloud. Doing that helps teams work together better, prevents clashing changes (thanks to state locking), and keeps things safer with encryption (env0.com, SquareOps).

They also spent time discussing Terraform modules. While modules definitely help with reusing code and keeping it tidy, they warned against getting too modular. The point is to package up meaningful blocks not break out every tiny piece. Having a clear architectural diagram early on can really help you avoid things like circular dependencies (Attract Group).
When it came to workflows, the suggestion was a clean, step-by-step pipeline: run checks first, then plan, apply (but with manual approval), and keep an eye out for drift. Tools like pre-commit hooks and sticking to short-lived branches can help smooth things out and avoid messy merge issues.

They also brought up some security tips. Since the Terraform state file is plain JSON, storing secrets in it is risky. So, encrypting it both at rest and in transit is a must (SquareOps).

The session wrapped up with a live demo that tied all of this together and showed how applying these ideas leads to better, safer infrastructure management in real projects.

Hackathon Hack the Future

On November 19 I got the chance to be part of Hack The Future, a yearly hackathon that's all about mixing tech, creativity and a good dose of collaboration. I teamed up with Harman Preet Singh my fellow AI student from Howest and together we tackled a challenge at the Flanders Meeting & Convention Center yeah wich is right next to the zoo of Antwerp.

We jumped into a problem proposed by EpicData called "Planet Hunters: In Search of a New Home Planet." After some coffee and a quick briefing we were sent off to our rooms and yeah that's when things got real.

The objective I found very interesting. Imagine Earth becomes unhabitable where do we go? Using machine learning and data science we had to dig through exoplanet data and try to figure out which ones might just might be habitable. What made it even more exciting was that we got to pick our own dataset. We didn't want anything pre-filtered so we dove into NASA's official Exoplanet Archive.

We started off with data exploration then focused on narrowing down features stuff like surface temperature how far the planet is from its star atmospheric data if available. We built visualizations and got a model running in Python using Jupyter Notebooks. We used ydata_profiling PyTorch and added in some of our own visuals too. After testing and tweaking we ended up with 13 exoplanets that looked promising.

Later in the day we shared our work in a final presentation. There were two types of votes one from a jury and one public vote. That second one was tricky since most people were from other schools so winning that one already showed to be a little hard. So at the end no trophy but we were still happy with the results we made.

If there's one thing that stuck with me? It's this: having good results isn't enough. You've got to explain it well. Our model did its job but I think we could've communicated the "why this works and how" part a bit better. Definitely something to work on next time.

The parts beside the project itself? Very nice. Good lunch in the lunch time also the chance to go into the zoo and a cozy wrap-up with champagne beer or soda. The vibe was relaxed but still competitive.

Hack The Future 2024 was packed. I didn't come out with groundbreaking new AI stuff but I left with more confidence in my teamwork great learned lessons and also left carrying a great memory.

Tech and Meet: Cyber Command Belgium

On November 12 2024 I went to a Tech&Meet session at the Howest campus where Cyber Command Belgium was the main focus. It was one of those events that really pulls back the curtain this one gave us a behind-the-scenes look at how Belgium is stepping up its game in dealing with cyber threats. What stood out right away was how cyberspace is now treated just like land sea air and even space. A full-on operational domain. Honestly that shift says a lot about how modern warfare's moved way past borders and boots on the ground. Now it's also about protecting digital systems and the stuff we rely on daily.

CyCOM which officially started in 2022 under the SGRS (Service Générale du Renseignement et de la Sécurité) is the unit behind all this. Their mission? Defend Belgium in the digital space. That covers a lot like making defence systems cyber ready running ops to block threats and teaming up with industry partners schools and other nations. Something I didn't expect was how diverse the team structure is. You can join full time part time as a reservist or even as a civilian expert. Sounds flexible but the entry process? Super strict. There's a full security screening that digs into your background family online activity everything. Understandable though considering the sensitivity of the work.

While I'm not sure a military path is for me personally I left with a lot of respect for what CyCOM is doing. Their scope goes way beyond just internal defence. They're fighting cybercrime sabotage espionage and even digital manipulation. And that affects all of us not just the military.

One thing that I really liked to hear was how seriously they take education. Their partnership with Howest and other schools shows they're not just defending but that they are building the future workforce too. It's great to see defence and education getting on the same page.

If anything the biggest thing I walked away with was this: cybersecurity isn't just an IT job anymore. It's national security. Whether you're a student already working in tech or even just interested in the field there's a role to play.

All in all a really eye-opening evening. It reminded me that today's battlefields aren't always physical and that Belgium's cyber defence strategy is already well underway.

Tech and Meet - Base Row

On October 22 2024 I attended a Tech&Meet session at Howest Campus Brugge that dove into how Baserow is shaking things up in the low code world. The talk was led by Frederik Duchi a product specialist in low code platforms and he walked us through what Baserow can do with a good mix of explanation and a live demo that really helped connect the dots.

Low code and no code tools are all about making app development faster and more accessible cutting down on the need for heavy coding skills. Baserow fits right into that space. It's open source and looks very easy to work with. You can view your data in different ways like Kanban calendar forms and hook into tools like Zapier or Make without much problems.

What sets Baserow apart is how much control it gives users. Unlike closed systems you can self host it dig into the code and stay compliant with things like GDPR or HIPAA. There's role based access audit logs and it can handle huge data sets with millions of rows. People are already using it for things like CRMs HR tools project tracking or just organizing data at scale.

But what stuck with me most was the open source philosophy behind it. Baserow really leans into data ownership and flexibility no vendor lock in no hidden traps. And during the demo Frederik made it clear how quickly you could turn complex data into something collaborative and useful without touching code.

If you're looking for ways to build tools faster and with fewer limits Baserow's definitely worth keeping an eye on.