Thirty years in technology, and I still had to start over.

I spent three decades building and running systems inside organizations — the kind of work where nobody claps when it goes right and everybody calls when it goes wrong. SharePoint, Microsoft 365, infrastructure, the long unglamorous middle of enterprise technology. I was good at it. I was also, by 2023, watching a wave come in that made a lot of that experience feel suddenly negotiable.

Here's the part I didn't expect: being a beginner again was the best thing that happened to my career. Not because it was comfortable — it absolutely was not — but because it put me back on the same side of the table as everyone I now teach.

What I actually do

I test things. Publicly, on purpose, with the failures left in. Then I teach the part that survived. Then, when a business needs it built rather than explained, I build it. That's three activities and they feed each other, which is why there are three doors on the front page instead of one.

The through-line is a refusal to oversell. There is an enormous amount of money right now in telling people that AI will do everything, and a smaller, more honest living in telling them exactly which parts it does well. I'd rather have the second one.

Who I'm for

People who were told, implicitly or otherwise, that this train left without them. Career changers in their fifties. Small business owners who can't afford a consultant and shouldn't need one for this. Grandparents who want to understand what their grandkids are using. Organizations who have paid for Copilot licenses that nobody opens.

If you want cutting-edge model research, I'm the wrong guy and I'll tell you who's the right one. If you want to know what actually works on a Tuesday with the tools you already have — pull up a chair.