Your platform problem probably isn't a tooling problem.
I'm Tomasz. I spent 20 years building platforms (Kubernetes, GitOps, CI/CD) for banks, telecoms and global enterprises. Now I don't build them. I sit with engineering leaders before the big calls and tell them what I'd do and why, without a vendor or an implementation team behind it.
Who this is for
- You're about to buy a platform product, and you suspect the real problem is somewhere else.
- Developers still open tickets to get a database or a new environment. Self-service has been "next quarter" for two years.
- Your AI pilots work on a laptop and stall the moment they meet your production platform.
- You just took over engineering at a new company and need to know quickly what's broken and what can wait.
Google and AI have the knowledge. I have the scar tissue.
— Tomasz CholewaHow I work
Diagnostic session
A fixed-price, 90-minute conversation focused on one real decision. Written summary within three business days. No pitch at the end.
Learn more →Advisory retainer
A monthly call, plus async access when something comes up in between. I advise; your team builds.
Learn more →My position
I've watched consulting firms and vendors sell things that didn't deliver what they promised, with recommendations shaped by whichever partnership paid the bills.
I don't implement and I don't resell anything. So when I tell you something doesn't fit your situation, I have no reason to say otherwise.
Most platform adoption problems I see come down to friction. When using the platform is harder than going around it, people go around it. More tooling rarely helps. Removing tickets and manual approvals usually does.
AI made writing a Kubernetes manifest or a pipeline nearly free. What's still expensive is deciding what to build and living with the architecture afterwards.
Most platforms running today weren't designed for agents. If yours still depends on tickets, manual approvals and knowledge in someone's head, AI agents will get stuck in the same queue as your developers.
Featured thinking
What is Continuous Data Recovery TestingOnce you follow the practices that I preach about defining and maintaining your whole platform with code, you’re simplifying and speeding up the recovery process. It’s a huge gain for an organization that prepares for the worst. It’s part of the “Design for Failure” principle that has been a part of best practices within the platform architects’ community.
7 lessons from writing AI agents for platformsI decided to bet on AI agents and started writing them a long time ago. There are many reasons, but the main one is pure laziness. I’ve always wanted to automate things, and that’s why I became a DevOps engineer in the first place.
5 Challenges of Using AI for Platform EngineeringAI is transforming software development, but platform engineering presents unique challenges. While LLMs excel at writing application code, they struggle with platform management tasks. Here are the 5 key challenges and why platform engineering needs a different approach to AI.
Selected clients I've worked with
"Tomasz has played a crucial role in revolutionizing our software delivery processes. His deep insights and expertise have been instrumental in helping us achieve our objectives and embrace the Cloud Native approach to modern application development."
"His insights and expertise on the subject were very helpful in understanding even the most complex of the inner workings of Kubernetes and Cloud Native projects. The program was specifically tailored to the company's environment and aligned with the tools and practices used in the participants' daily work."
Start with one conversation.
The diagnostic session is a fixed-price, no-obligation first step. If it's not the right fit, I'll tell you.
Book a diagnostic call