Getting an AI pilot into production
Most AI pilots work on a sample and stall before the business relies on them. What has to be true for one to run in the daily work, and who has to own it.
Guides
How I approach the decisions clients bring me: where AI money goes and how to get AI into production, how to step in as an interim COO, and how a strategic CFO plans the money. Written from the work, mostly in healthcare and life sciences, for any company facing the same decisions.
Forward-deployed AI
Most AI pilots work on a sample and stall before the business relies on them. What has to be true for one to run in the daily work, and who has to own it.
AI can take real cost out of finance and operations work. Most of the savings come from what you clean up before the AI goes in, and most of the damage comes from skipping that step.
Automating tasks inside a function that stays the same size saves a little. Moving the function itself onto AI changes how many people a company needs. How I decide what moves, build it and keep people where they matter.
Most companies don't have an AI strategy problem. They have dozens of AI projects nobody can compare. How to get them into one view, score them the same way and decide what to fund, fix or stop.
Interim and fractional COO
An interim COO is hired because something changed: a leader left, the company outgrew how it runs, or a deal has to be carried through. What the first three months should look like, and what to expect at each step.
How to evaluate a new business line, write the plan that decides it, and build the division that runs it. Drawn from building several businesses inside Optum, including one built from nothing into a major operation.
When costs have outrun the business, cutting everything a little is the slowest way out. How to find what makes money, protect cash, cut deliberately and keep the operation running while you do it.
Fractional strategic CFO
Built from drivers, with scenarios, a runway in months and dates, and a capital plan tied to milestones: how to build a model that drives decisions instead of decorating the board deck.
Most board updates report what happened. The useful ones say what changed, what it means for the plan and what decision is needed. A structure that works for boards, owners and investors.
Most acquisitions that disappoint were decided on a story and integrated as an afterthought. How to test the target, build a business case you can hold the deal to, and plan the integration before you sign.
Start with a conversation
You describe the situation: a workflow, an operation that needs a leader, or a plan that needs a CFO. I tell you what I would do first.