About

I am an experienced freelance Linux systems engineer and open-source enthusiast with 31 years of hands-on experience in managing and optimizing Linux environments, on-prem as well as in AWS and GCP.

My expertise spans both Red Hat Enterprise Linux (RHEL) and Debian-based distributions, making me proficient in a wide range of Linux systems.

I have worked as a sysadmin, Oracle DBA, DevSecOps engineer and Google Cloud Architect at multiple MSP's and as a freelancer for over a decade. The broad knowledge I have gathered in IT, led to the decision to pursue my passion professionally, which is finding creative solutions to complex problems using open source components.

-- Mark

Open Source

I'm a huge proponent of open source software. Having worked with big-tech products for decades, I started to notice a pattern of dependency without the required control to adjust for ever more rapidly changing infrastructure needs. This lack of digital sovereignty can have huge implications on the ability of a company to build and maintain a sustainable and resilient infrastructure.

In my opinion, a company or government entity should strive for digital sovereignty, which means they should have the ability to control their own digital infrastructure, technologies, data, and online processes without an external dependency on foreign entities or large technology companies.

The best and most efficient way to achieve that, in my opinion, is to move away from big tech and towards open source solutions. Several EU governments have already begun this process by exchanging Microsoft software and services for open source solutions.

This movement gained momentum when the European Commission appointed Henna Virkkunen as the first executive vice president for tech sovereignty, security, and democracy in 2024.

Artificial intelligence

Artificial intelligence is taking the IT industry by storm, there is no denying that. I personally think it's going to be the next dot com bubble. The reason I say this is because when you approach AI from an economic perspective, the unit economics just don't add up. Historically, unit economics dictate that the price of "a unit" should go down over time, as production cost decreases and demand increases, but this is not the case with AI.

The definition of "unit" and the cost structure are uniquely shaped by how AI services operate. The unit varies by business model, but I think the "unit" can be ultimately distilled down to a "token", which is the smallest granularity. Instead of going down, the price keeps going up, which in itself is an indicator of a bubble.

Add to that, only 28% of AI projects succeed to meet ROI according to Gartner, that's pretty abismal in my opinion.

In my personal experience, where I ask AI for solutions to coding problems (to which I already know the answer) I usually receive an answer that either compiles and runs but is insecure, or it won't compile at all due to hallucination. Companies are opening themselves up to scenarios of inefficiency at best or getting hacked at worst. I guess time will tell.

Penetration testing

Over the last two years I have expanded my cyber security knowledge through the penetration tester career path at HtB. This supplemented my skillset with that of a potential attacker perspective and was a very insightful experience.

If I can take away one thing from these newly gained perspectives, it's that the Unix philosophy is more important now than it has ever been, even though parts of the market seem to often be moving in a different, perhaps even opposite direction.