What's next, and why it matters.
Notes on new technology: AI, the tools changing how we work and create, and what I learn by trying them myself.
Platform engineering, DevOps, SRE: what's the actual difference?
Three job titles, one confusing overlap. What each one really does, and which one your team needs.
Golden paths: standards that don't slow anyone down
How to make the right way the easy way, with escape hatches for the edge cases.
Are we actually getting faster? Measuring DevOps with DORA
Five numbers that tell you how your delivery is really going, and the vanity metrics to ignore.
Portals, CLIs or Git: what developers really use
A shiny portal is nice. But is it what your developers actually open every day?
Less to think about: reducing cognitive load for developers
Every YAML file you ask a developer to understand has a cost. How to make it smaller.
Monorepo or many repos?
One repo to rule them all, or one per service. The honest trade-offs for CI and ownership.
MCP: plugging AI into your infrastructure
The Model Context Protocol lets AI talk to Kubernetes, Terraform and GitHub. Powerful, and risky.
AI writes the code. Who checks it?
AI-generated code brings new risks, including packages that don't exist. How to catch them in CI.
Too many alerts: can AI cut the noise?
When everything alerts, nothing does. A look at AIOps and smarter alerting.
GPUs for DevOps people
GPUs are now part of the infrastructure. What a DevOps engineer needs to know about them.
Will AI take my DevOps job? One year later
A year ago I made predictions. Time to check which ones came true.
What's new in Kubernetes, and what I'm upgrading for
The release notes are long. These are the changes that matter for day-to-day work.
Goodbye ingress-nginx, hello Gateway API
The most popular way into a Kubernetes cluster is retired. What that means, what Gateway API is, and how I'd move over without a weekend of downtime.
Helm 4: what changed
The first major Helm release in years. What's different, and whether your charts still work.
More than one cluster
Why teams end up with many clusters, and how to keep them from becoming many problems.
Databases on Kubernetes: yes or no?
Running state on Kubernetes is possible now. Whether you should is a different question.
Argo CD or Flux?
Both do GitOps well. The differences that matter when you pick one.
GitHub Actions, GitLab CI or Jenkins in 2026?
Is Jenkins still worth it? An honest comparison by team size.
Terraform or OpenTofu?
A license change, a fork, an acquisition. Which one I'd pick today, and why.
Is Ansible still worth learning?
Terraform and Kubernetes took the spotlight. Where Ansible still wins.
Where does your software come from? SBOM, SLSA and signing
Supply chain attacks are everywhere. The three ideas that help you trust what you ship.
The EU Cyber Resilience Act and DevOps
New EU rules for software security started applying in 2026. What teams need to do.
No more static passwords between services
Workload identity and SPIFFE: letting services prove who they are.
When npm turns against you
Worms, typosquats and hijacked maintainers. How I protect builds from malicious packages.
Quantum computers and today's encryption
Quantum computers can't break encryption yet. Why DevOps teams should prepare now anyway.
SLOs: deciding how reliable is reliable enough
100% uptime is the wrong goal. How error budgets help teams decide.
eBPF: seeing inside Linux without touching your code
The kernel technology behind modern networking, security and observability tools.
Alerts that deserve to wake you up
If an alert doesn't need action, it shouldn't page anyone.
When things break: incidents and blameless postmortems
The roles, the messages and the postmortem template I use.
Multi-cloud: an honest look
The real reasons for multi-cloud, and the hidden costs nobody mentions.
European cloud and digital sovereignty
Why "where is my data?" became a boardroom question in Europe, and what the options are.
Leaving the cloud: when it makes sense
Some companies saved millions by moving back to their own servers. Could you?
Serverless or containers?
Lambda, Cloud Run or Kubernetes: which fits which workload.
Service mesh without sidecars
Istio ambient, Linkerd, Cilium. Do you even need a mesh?
WebAssembly on the server: hype or future?
Wasm escaped the browser. Is it the next container?
Should DevOps engineers learn Go?
Kubernetes is written in Go, and so is almost every tool around it. Where to start.
Docker, Podman, containerd: what's the difference?
What actually runs your containers, and the Docker Desktop alternatives.
How I'd learn DevOps if I started today
A step-by-step roadmap, with a project for each step.
DevOps interview questions I'd ask
The questions, and what a good answer sounds like.
Which DevOps certifications are worth it?
CKA, AWS, Terraform: which ones employers actually care about.
Portfolio projects that get you hired
Fifteen projects, from beginner to advanced, that show real skills.
From sysadmin to DevOps
What carries over, what's missing, and how to make the move.
On call without burning out
Being on call shouldn't cost your sleep or your health.
Why tools alone don't make DevOps
You can buy every tool and still have silos. Culture is the hard part.
Green DevOps: the carbon cost of our servers
Efficient infrastructure saves money and emissions. How to measure both.
DevOps in 2026: what changed this year
AI agents, platforms, new security laws. My look back at the year, and what's next.
Goodbye passwords
Passkeys explained simply: what they are, why phishing stops working, and how to switch your email over in five minutes.
Will AI take my DevOps job?
An honest view from inside: what AI already does in pipelines and on-call, what it can't, and how the job is shifting.
Proving a photo is real
AI can fake a photo in seconds. How Content Credentials let real photos prove where they came from, and the catch.
Your next car is a computer on wheels
Over-the-air updates, features sold as subscriptions, and what a car person wants from a software-defined car.
AI & MLOps, from the ground up
Notes from the very start of my 24-week journey into building and running real machine-learning systems.
The camera is software now
How computational photography changed what a "good camera" means, and why my phone often beats my DSLR.
AI in everyday apps
The AI features I actually use, the ones I've switched off, and how I decide.