POV: AI runs on the layers underneath it
Most companies I meet want to buy AI before they can tell me how their business actually runs. I understand why. The demos are good. But AI does not run on enthusiasm. It runs on t
← Home
Writing
The foundation is everything (AI Operating Model)
The foundation is everything. Miss it and you're capped, no matter how good your tools look. I keep seeing two versions of the same trap. In finance: a business looks decision-read
AI does not raise a multiple. An operating model does.
I recently spent time in a room full of private equity operators, and the conversation kept landing in the same place. Everyone has bought AI. Almost nobody has turned it into ente
Pick one problem, prove it worked
A first operations assessment usually surfaces dozens of problems. The instinct is to fix them all at once. That is how nothing gets fixed. The hard part is never the list of probl
You can't automate what you can't describe
Your AI pilot didn't fail because the model was bad. It failed because nobody could describe how the work actually happens. Most businesses run on processes that live entirely in o
Almost everyone adopted AI, almost nobody was transformed by it
I keep coming back to one thing: almost everyone has adopted AI, and almost nobody has been transformed by it. The tools are the easy part. What separates the few who cross from th
The AI is 10 percent, the operating layer is 90
I keep seeing the same ratio, and it is backwards. Everyone treats the AI model as the whole answer. In a real business it is maybe ten percent of it. The other ninety percent is t
Certainty before commitment
Most transformation projects ask you to bet the budget before you know a single thing. The teams who demand proof before they commit quietly outrun the ones who gamble. And I have
Audit operations like you audit your books
Most businesses run their operations on a system they would never accept from their accountant. The ones who audit the work itself, the way they audit their books, are the ones AI