Speaking

Sessions written to be given. They are not a calendar of past events. If one fits a room you are hosting, the conversation starts on LinkedIn.

The queue was not a hiring problem

A 25-minute talk on why support load, delayed answers and “we need more people” are often a system defect.

The room usually arrives with a headcount request. The useful question is different: which decisions return to the same person, which answers are rebuilt every week, and which system of record nobody trusts.

I walk through a support queue that looked like a staffing gap and was, in fact, a missing operating rule: no owner for the answer, no definition of done, and no path back into the product.

Audience: founders, COOs and engineering leaders of companies between roughly 30 and 300 people. Format: talk plus a short working diagnostic the room can apply the same week.

The company outgrew its operating system

How growth turns tools into noise, and what a leadership team can install instead of another platform.

Most “digital transformation” language starts with a tool. This session starts with the fracture: meetings that create meetings, reports nobody believes, and a founder who is still the integration layer.

The second half is concrete. Four moves — see the system, name the friction, choose a sequence, install a cadence — and the artifact each one leaves behind.

Useful for offsites where the leadership team is about to buy an ERP, an AI program or a new head of technology and has not agreed what problem those things are for.

AI governance for operators

A practical session on where models may draft work, and where the company still needs a human decision.

This is not a model tour. It is a governance conversation for people who run a P&L, a plant, a support org or an engineering group.

We separate three uses: drafting, classification and decision. Drafting can move fast. Classification needs a sample and an owner. Decisions that spend money, change a customer or ship risk stay with a named person.

The leave-behind is a one-page allowance: what the company will try this quarter, what it will not automate, and how it will know the experiment worked.

Book an executive session