Nowadays everyone talks about the CS degree, like it's worthless. Like no matter what they do, they won't find a job and are hopelessly meant to be replaced by AI. But no one wants to do the simple things to actually become job ready on a day to day basis. Today I am gonna give you the 3 skills, you need to learn to get a bullet proof IT-Job in 2026 and I am gonna tell you how to learn them step by step.
Your CS Degree is not worthless.
Before we start, I wanna get this out of the way, cause I see so many people online saying the CS degree is dead and you should just drop out. This is plain wrong and most of those people didn't even study CS.
Your degree gives you three things most self taught devs don't have:
- Academic leverage. In Germany (and most of Europe) employers still look at your degree. It gets you in the door, it gets you interviews, it makes HR take you serious. Like it or not, this is how the market works right now.
- Connections. You sit in lectures with 200 other people, who will later work at companies you wanna work at. Your professors know people in the industry. Your fellow students will refer you later. This network is worth more than any course you could buy.
- The basics. Yes the coding you learn at uni is far from what the job wants, but you learn how to think. You learn algorithms, data structures, databases, how a computer actually works. This is the foundation everything else sits on.
So your degree is not the problem. The problem is that most students only do the degree and think they are done. They are not. The degree is like 30% of what you need, the other 70% you gotta do on your own. And that's what this guide is about.
Figuring out where to go
There are different paths you can take in the IT/CS field, but the three skills I am gonna share with you are the three basic skills, you will need in nearly every Software related field you can go to. If it is system administration, Software Engineering, Cloud Engineering, DevOps Engineering, something AI related. Those 3 skills are the basics. This guide won't tell you what job to go for, but I will tell you the skill set, that currently seems to be very high in demand and basically almost irreplaceable by AI and will allow you to get any job you want later on.
Those skills are: Infrastructure + Coding + AI.
Why those three skills are so hard to replace and so in demand is to much to cover for this article. Believe me I have heavily researched this topic and I work with experts in the current field, that will tell you (and are telling me) exactly the same stuff. So, when it comes to what to learn, just follow my guide for 6 months to 1 year and then rethink, in which direction you wanna use that skill.
Explaining those three skills and a few misconceptions.
On Coding. Well when most people wanna get a tech job they mean this. They mean coding, so writing code and developing applications(apps/software). You should have basically learned this and gotten some coding skills through university by now. But there are two problems here: a) The coding you learn at university is far from what you will actually need at a job b) AI and AI Agents are on their way to almost fully replace Software Engineers that only code.
Those two problems create three necessities:
a) learn what type of coding skills are needed in the current market b) learn and understand them, so you can practically use AI to do this for you c) learn the skills that are actually harder for AI to replace and in demand currently and in the future.
Those harder to replace skills are: Infrastructure + AI
On Infrastructure. When you have applications the software gotta run somewhere for the user to use this. Like your Game running on your pc, which is providing storage and compute power for it to work. Your game applications run on many different computers which we call servers.
Now your application used to run directly on the Server back in the day (or rather directly on the OS of the server), but overtime we developed other Software on which the application can run on. So your game not running directly on the operating system, but on another software that is a layer between the application(game) and the OS (Windows or Linux). This Layer basically makes everything easier, faster and more efficient. The latest software that is used here is Kubernetes.
So, when I say learning Infrastructure, I mean three things:
- provisioning and maintaining the servers your app runs on (mostly in the Cloud)
- Managing the Software your apps run on and running the apps on that software (mostly Kubernetes for production)
- Automating a lot of this as much as possible.
Just for you to understand why learning Kubernetes is so cool, here is a resource:
- The current pay of Kuberntes Jobs: https://kube.careers
So combining those skills, will make you being replaced less likely. Since you understand the logic behind the code(what AI does for you), know how to debug and have the Infrastructure skillset, which AI currently can't and won't for the near future. Also Kubernetes is in high demand in the last few years, with the demand growing and growing.
So now combining those skills is very important, to a) be the one that understands how the software is made and b) just do the infrastructure work that can't be replaced.
On AI: When it comes to AI to not over complicate it for you, this mostly means learning how to use AI productively and with leverage, so you can do all of this better and at a higher productivity than your co workers.
Now I hope this is clear, from here on out I will explain you how to get a job.
A nice little study split
For maximum productivity you need to stop binge watching courses and do a healthy mix of Theory for new information and building to actually apply it. Therefore you will split your tech study day into two sessions everyday:
- Theory
- Building
So the theory stuff is about 30 mins - 1 hour at the beginning of the day, where I read books or watch courses and force myself to actually understand what I am doing and learn new stuff. To be honest I don't like this, I don't believe this to be fun, but this is a must have to grow.
Now when it comes to building, here just at the end of the day, when I finished my work I just build projects, selfhost applications, work with AI and all of this stuff to get practical experience for 30 mins - 1 hr depending on how much time I have.
This should be part of your day, building and studying, this is your foundation for growing in the long run and the stuff you can show to employers when they ask you what you can do, for good reference I would recommend you to make a Github profile and put your projects there.
Doing this is the only way to get something close to work experience and become interesting for recruiters. They will see you have a hands on mentality, love what you are doing, are dedicated and this is basically proof for them and can compensate for work experience.
The general split for theory and building is the basis of this and the study structure I just presented is a very good one, but it depends on the time you have. What you can do instead:
- Alternate one day theory one practice
- Do 30 mins - 1 hr theory, rest of my freetime practice
- Integrate practice into theory: prolonged theory phases, followed by building projects
What you do is yours, but generally it's important that you try to open your laptop everyday and learn a bit to build momentum.
The roadmap
I already told you that I focus my learning session around 3 topics: Infrastructure + Coding + AI. I gotta be honest here, If you are new to the game, I would recommend you to just completely copy me. Don't do anything else, don't fall for shiny object syndrome, just follow my plan for about 1 year and you will see the result.
So for theory I would recommend you the following:
- Go here: https://www.skool.com/linux
- Make an account if you need
- Join the Community
- Watch through it and finish their courses -> especially the Linux Course
- Then go to Learn2cloud.io
- make an account
- follow the curriculum until you arrive at Kubernetes (you can skip the Linux Book)
- When you arrive Kubernetes:
- Buy this: course (wait for it to be discounted)
- Finish it
- Continue with learn2cloud and finish it
- Go here: https://anthropic.skilljar.com/
- learn claude code and become good at using it
This was our Basis, it should take about 6 months to 1 year to finish, but will give you if you dedicate yourself to it, more skillset than 99% of CS Students. You will know about latest AI Tools, DevOps technologies and Coding. Like this is it, from here on you would need to do something I call "Becoming professional" to grow even more, but this is for another video, so subscribe to get notified.
When it comes to building:
As I told you we will split our learning sessions into two, so here's what to do in the building sessions:
- Before you finished the Free Linux Course:
- Don't work on projects in your free time yet, just work through the course, to get a basic understanding.
- Then:
- Install Arch Linux on an old piece of hardware or the rasberry pi, you will have already configured in the Free Linux Course:
- once with this course: https://www.youtube.com/watch?v=FxeriGuJKTM
- second time only with notes, you take whilst doing the course -> this will give you a very deep understanding of Linux and will make you feel so helpless in between.
- When you have Arch Linux installed:
- Delete it and install Ubuntu
- Then Go Here:
- https://overthewire.org/wargames/bandit/
- do those challenges one by one, to become more Linux fluent (I would recommend you to get to like Level 15) After this you can stick with learn2cloud and do their projects, very simple until you have reached the Kubernetes section.
- Install Arch Linux on an old piece of hardware or the rasberry pi, you will have already configured in the Free Linux Course:
Building a K8s Homelab
Once you have finished the Kubernetes course I have recommended I want you to use your old laptop or your rasberry pi, on which you installed ubuntu and build a Kubernetes homelab. Because now you are at another level. After the Kubernetes Masterclass you basically almost have finished Learn2cloud(I would recommend you to finish it completely). And now you have quite a good basis.
As I told you, after Learn2cloud I would recommend something, I call "becoming professional". It's a structure I have developed for myself, to level up in my career. And this project is actually a part of this structure. Now I gotta admit I didn't have this structure 2 years ago, that's why I a few weeks ago needed to revisit a bunch of concepts, going through some Learn2cloud and Free Linux course concepts again and again, as I have missed basic theory. So I would recommend you to not skip to this part too early.
What we are gonna do is: Build a production grade K8s cluster using Talos Linux. If you don't know what this means no problem, you will understand this, when you follow the rest of the curriculum up to this point.
The homelab is basically like a production grade Project, companies can see that proofs your skill to them, because it's something very hard to manage and needs a huge amount of skill to build. Since it's to hard to manage for a beginner, our end goal is to have this described Production grade system, but for the beginning I would advice you to do the following once you reached the step of building a homelab.
- Buy a raspberry pi or use an old laptop
- Install Ubuntu as a Server on it
- Connect it to your work laptop over Tailscale
- Install K3s on it
- Start deploying stuff
It's very simple to do and to do it, I would recommend you to just give those steps to any AI you want, then specify, like tell him what laptop you have and so on and do it together with him, try to understand every step.
Implementing AI
Next up is integrating AI here, but this is for the future and a next step of the "becoming professional" Framework. This would outweigh this article. Now what I can tell you to look at here more deeply is:
- How do I integrate Claude Code in my workflow?
- Am I interested in AIOps or MLOps?
- What projects can I build with AI that actually make sense?
Generally you don't wanna overly depend on AI, but rather learn how to work with the latest AI technologies and how you can possibly pursue a career in the AI field.
This alone is enough for like the next 1-1.5 years and will give you a huge headstart.
Building a career.
Now to build a career, you will need more than just tech skills. You will need soft skills, the ability to present yourself, understanding of the market and so on. So for this I would recommend you to follow my channel as I will give you more and more of the Frameworks I use to build a good position at my company and slowly design my own career path.