Hey Reader,

​

Three months ago, I wrote an article. I said that if you want an infrastructure job in a Fortune 500 company, you should learn Kubernetes. I also claimed that Kubernetes would become the AI platform. Hundreds of people said I was wrong.

This week, the CNCF dropped its annual survey with the latest data. Let me show you what they found:

Kubernetes Won

Most of the angry commenters on that article gave me the usual push-back:

• Almost no company needs Kubernetes
• It is overkill for everything
• It is too hard to learn

Here’s what the latest data says:

• 98% of organizations use cloud native techniques
• 82% of container users run Kubernetes in production (up from 66% in 2023)
• Container usage in production apps increased from 41% to 56%

This isn’t a fad. Kubernetes is over 10 years old now. It has won the infrastructure game.

Do not sleep on this. If you want an infrastructure or DevOps job, this is where the money is.

The talent shortage won't last forever. Every month, more engineers catch on. Every month, the gap closes.

I'm opening 10 spots for February. Applications close when they are filled.

APPLY NOW → CLICK HERE​

​

​

​

​

Kubernetes Is The De Facto AI Platform Now

I’ve been saying this for years, but now I have the CNCF proof:

Many engineers I talk to don’t know that the majority of AI workloads run on Kubernetes. ChatGPT, Claude, and Grok are all built, trained, and run on Kubernetes.

This is also why there are over 120,000 open jobs right now with an average minimum salary of $140K.​

Every major wave, containers, microservices, virtual machines (KubeVirt), and now AI, lands on Kubernetes. The platform keeps winning. The skill keeps compounding.

I went all-in on Kubernetes 5 years ago. Now my inbox is full of recruiters every single day.

The Skills Gap

Despite massive adoption, companies can’t find people.

The CNCF AI Talent Report from last year backs this up. Over half of organizations say they don’t have enough staff in cloud and platform engineering roles. Also, 68% are missing personnel in AI and ML operations.

Because of the lack of people, the actual AI work is also slowed down. Only 7% of companies deploy daily.

Fifty-two percent of companies with AI workloads don’t train models. They use Kubernetes to serve the models only.

In other words: they haven't even started training their own models yet, because they don't have the people to do it.

This is where there is a huge opportunity for infrastructure engineers.

These companies need help with the essential infrastructure for their model pipelines. This includes strong CI/CD, monitoring, and resource optimization.

People with Kubernetes skills will become the new dusty system administrators.

These roles offer stable jobs that provide high salaries, but few people are aware of them.

​

Here’s the main point: we’re still at the very start of AI infrastructure, but Kubernetes has already come out on top.

There’s a lot of work to be done in the next 10 years to make this mature.

And that is why they will pay you $200K a year.​

​

People Are Not Learning It

The report highlights the key challenges organizations face when running containers on Kubernetes.

• Cultural changes with the development team
• Lack of training

Kubernetes is 10 years old now; Docker has been around for 13 years.

Still, in 2026, companies say that “cultural changes” is their biggest challenge.

Here’s what that means:

There are armies of reluctant software engineers who don’t want to change their ways. They want to stick to what they know, and they resist adopting new technologies. Even if people consider that technology “boring” at this point.

I know someone like this who refuses to use Git because he’s used to TFVC.

Containers and Kubernetes frighten them, and thus they will not take it on.

This is where your moat is. These skills will set you apart and make you a much more attractive candidate, even as a junior.

I experienced this firsthand while I was working as a Junior DevOps Engineer five years ago. The day I discovered Kubernetes, I fell completely in love with it. I saw that this would be the primary method for deploying software in the future.

I used my evenings and weekends to learn Kubernetes. My manager and mentors discouraged me, but I kept going. Six months later I doubled my salary by becoming a Kubernetes Consultant.

What should you Do?

It’s simple. Learn Kubernetes. Properly. With hands-on projects, not browser labs.​

You can’t learn Kubernetes without a solid grasp of Linux and containers. If you have not mastered those yet, start there.

If you’re already skilled in Kubernetes, these are the tools to focus on this year (based on the report):

• KServe: used for model inference
• Kubeflow: used for model training and MLOps pipelines

Average outdated engineers spend their time debating which AI model is best. Engineers who know how to deploy them and run them are being hired.

The window is open. But windows close. Adoption is accelerating. The talent shortage won’t last forever.

3 months ago, I told you Kubernetes was the AI platform. The data has confirmed it.

Are you going to take this opportunity?

Let's get after it.

Mischa.

​

P.S. Five years ago, I was mass-applying to jobs nobody wanted me for. Six months after I went all-in on Kubernetes, I doubled my salary. Ever since recruiters are chasing me every day. That path is open to you right now. But the window won't stay open forever.

​APPLY NOW BEFORE SPOTS RUN OUT​

​

This student applied and changed his life:

​

​