About

The center of it all is AIM-TO-PACE: a prompt framework. By now one that actually earns the label — not yet another memorize-this prompt pattern, but a small prompt ecosystem.

And yes, it's about prompting. I use agents, skills, MCP, I vibe- and loop-code. But the prompt remains the machine code that isn't one. That's good and bad, simple and complicated at the same time. Getting started is easy, anyone can prompt — the pitfalls are there all the same. And a prompt can be so much more than what most people get out of it.

That's exactly where the path lies for me: experimenting and learning at the prompt, at the machine. Chain two or three prompts together and you're practicing orchestration live — and putting it to productive use before IT has even picked up the ticket for the next agent. You can dive as deep as you want. Or stay at the surface and still keep your finger on the pulse with one or two rules and tricks — instead of trusting a prompt generator that's always a step behind and only works in that one spot. Simplifying is good and right. But why slap a form on top and wall in something that's still so much in flux?

That's the core: platform- and version-independent, close to reality, honest — and for anyone, with whatever they have at hand. An internet connection and access to an LLM — a free one will do. Taking away the fear and the hesitation. Raw and yet simple. First I tried this for myself, then for friends and colleagues — now for everyone. I want everyone to be able to experience AI firsthand. Without crossing seven bridges only to find a sealed book at the end.

The framework itself is field-tested — if only in a small circle — and free to use under CC BY 4.0. And that's how it stays.

Around it sits whatever else might help: Pace Prepped Prompts, for everyone who doesn't want to pay a discipline tax on their magic prompts. A catalogue of pathological LLM failure modes — they won't go away, no matter how good the models get. AI will always make mistakes. Once we've digested that, we can start dealing with it cleverly.

Then a few field notes — from my very personal point of view — on what worked and what didn't.

Rounded off with small tools I built for myself and saw no reason to keep to myself.

And there's more on the workbench.

Why all this? The gap between what the technology can do and what people actually get out of it grows wider every day. And most of the advice on the market comes from people who never had to roll anything out. I'd rather pass on the things that survived contact with reality and earned my personal seal of approval — at maximum risk of bias, of course. The IKEA effect in its purest form.

It started for fun: private tools, weekend experiments, a growing stack of things I'd figured out for myself. Then it outgrew that. At some point the framework, the meta-prompt, the ideas, the tools, and the experience had become something with edges: clearly defined, usable, mine. So now I'm self-employed on the side — a one-man company. Two years ago I would have dismissed that as completely absurd.

Will it work out, last, keep evolving? I don't know. Or rather: it already has worked out. I've learned more in the past months than I could have planned for. I'll keep going until it's over — and it doesn't look like it's over.

There's one thing I don't write about — on purpose.

Nothing from inside my day job. No internal cases, numbers, names, or systems — and nothing here is a statement by my employer. What's written here is mine and speaks only for me.

Everything on this site was created privately — evenings, weekends, on my own dime and my own infrastructure. AI is part of my day job in corporate IT, but only one part among everything else that running a platform takes. Most of what you find here grew on the side — it didn't just fall off the back of my job.

And sure: I'm one person, not two neatly separated selves, one private and one professional. But I do separate the content. Of course my work life has shaped how I look at these things — and vice versa. My employer knows about the side business and benefits from what I learn on my own time, rather more than the other way around. The lessons and insights travel. The details stay where they belong.