How I Used AI to Prepare for the CKA

I went into my CKA preparation thinking I’d be grinding through mock exams until I felt ready and confident. Instead, I ended up using AI to generate around 10 custom exams, targeting exactly the topics I struggled with. And honestly? It worked way better than I expected.

There’s a GitHub repository called CK-X (https://github.com/sailor-sh/CK-X) that lets you download and run mock exams for the CKAD, CKA, and CKS in a simulator that’s very similar to the real exam environment.

Unfortunately, at least for the CKA, there are only two mock exams available: one basic and one advanced. If you want more, you’ll have to use their website and pay for a bundle of additional exams. You could argue that the free repository is also a way to get people onto their website and encourage them to purchase more exams.

By that point, I’d already completed all the KodeKloud mock exams and both CK-X exams, but I still felt I needed more practice, especially in areas I wasn’t as confident in, such as NetworkPolicies and RBAC.

Then it hit me: I already had the CK-X repository, which I could use as a skeleton to generate more exams with Claude Code, tailored specifically to my needs!

Within a few hours, I’d generated two new exams that mimicked the CKA exam and focused on my weaker areas. Whenever I had doubts about a question, I could simply ask Claude Code for an explanation.

Sure, I had to tweak CLAUDE.md a bit, and some questions were confusing at first. But with a little patience and iteration, I got everything working smoothly. By the end, I’d generated around 10 mock exams!

I passed the CKA with 85%, and honestly, I think this was one of the best ways I could have used AI during my preparation.

The cool part isn’t that AI magically taught me Kubernetes. It’s that it helped me practice more, focus on what I was bad at, and get explanations whenever I got stuck.
Sometimes, you don’t need a better study resource. You just need to build one.