What we've learned, written plainly.
These are field notes from standing up AI and data capabilities — the arguments that win funding, the governance that helps you ship, and the unglamorous work underneath the models. Written for the person who just got handed the job.
Standing up an AI & data program
A working field guide for whoever just got handed the job of building an AI and data capability from nothing — not one model, but a program that creates value and an office that keeps creating it after you leave.
Talking about AI in dollars
Engineers talk accuracy. The CFO talks money. The gap between them is where most AI programs lose their funding — and here is the language that closes it.
Governance that helps you ship
Governance that only ever says no gets routed around — and then you have no governance at all. The kind that lasts is built into the gates, not bolted on before launch.
Why hub-and-spoke keeps winning
Centralize everything and the work stops fitting the business. Push it all into the units and you lose leverage and consistency. The hybrid wins for a reason worth understanding.
The pathfinder playbook
An operating model on paper is a hypothesis. Pathfinders are how you prove it with real work — and how ownership moves from us to you across a handful of live use cases.
The data behind autonomous agents
Agents plan, retrieve, and act on their own — and they fail badly on the data most enterprises have. What separates a demo from a workforce is the layer underneath.
Teach the insider
You won't hire your way to scale — the market is thin, and even a full roster wouldn't carry the domain knowledge that makes the work useful. The stronger move is often the person already sitting inside the business.
Want to talk through any of this against your own situation?
A first call is the fastest way to find out where you would start.