I’m working on a fascinating project that leverages AI to tackle Platform Engineering challenges.

It’s time to harness AI to make platform engineering more efficient. With more applications coming to our platforms, we must prepare for this influx using LLMs and AI Agents.

Why I’m building it

I’ve always been a passionate advocate and practitioner of the Everything-as-Code approach (formerly Infrastructure-as-Code). I ❤️ creating platforms managed entirely from code (a.k.a. GitOps) and have explored various tools to achieve this goal.
First, it was Kickstart, which I used to boot up machines via PXE and install Linux in an automated and unified way.
Then came the cloud, with CloudFormation, Terraform, Ansible, and Packer, enabling autoscaling and immutable infrastructure.
Finally, Docker appeared, followed soon by Kubernetes, revolutionizing infrastructure management and taking platform management to a new level!

But here’s the thing: we don’t need Terraform, OpenTofu, Kubernetes, Docker, Argo CD, Jenkins, or other fancy tools. What we truly need is an EASY way to run apps using Kubernetes as a foundation.

I know how it should work, which components to use, and it looks very promising! I’ve started building it to address common challenges in the Platform Engineering and DevOps fields.
Over the past 20 years, I’ve seen many companies struggle to provide a reliable, secure, and efficient platform for running their apps.

More expertise needed

One of the main reasons for these struggles is the shortage of DevOps experts with broad knowledge of complex infrastructure.
In many cases, AI outperforms humans in implementing solutions with code. And it continues to improve with more sophisticated LLMs under active development.

After months of learning and experimentation, I finally discovered how AI can address many challenges of maintaining platforms, particularly those based on Kubernetes.

AI works best with code

I’ve learned how LLMs work, and the conclusion is simple: they need text to understand context, connections, and intents.

The platform can be managed entirely from code, building and delivering apps with the help of AI.
To fully utilize AI for platform management, you must keep everything as code. This opens the door to significant opportunities for saving time, simplifying deployments, and enhancing security.

AI Agents managing platform via MCP

AI agents can use the code to improve and develop the platform.
MCP is used to enhance the platform by keeping all changes in code.

Kubernetes enables you to run apps in multiple regions, clouds, and providers at a low cost.
Code-based management enhanced with AI makes it a future-proof solution.

What next

Is it ready? NOT YET.
Do I know how to build it? YES.
Will it transform the way organizations manage their platforms? DEFINITELY!

Would you like to find out more? If so, please contact me directly. Let’s talk!