I build the systems that let a business run without me in the room.
Governance, SOPs, KPI systems, and the AI layer on top of them — designed and built personally, across three industries.
Eighteen years. Three industries. Departments built from zero.
The people were capable. The ambition was real. What was missing connected the two.
In practice, that meant the same rollout every time regardless of who ran it, the same information in front of everyone before a review started, and one set of numbers the whole business actually trusted.
Not another tool. Not another dashboard. Not another process. What was missing was one system connecting all of it — built once, so the business didn't depend on anyone's memory to run the same way twice.
Founders and leadership teams whose businesses have grown past what one person can hold together on their own.
If one of these is true right now, that's usually where to start.
Every new location becomes its own one-off project — nobody's run this playbook before, so it gets reinvented each time.
If expansion is getting harder as you scale, not easier — if the fifth location still takes as much of your time as the first one did — that's usually a missing playbook, not a people problem.
Read how this scaledThe business runs — but only because you're personally in most of the important rooms.
If decisions stall the moment you're unavailable, and partners or managers are waiting on your judgment for things that should be routine by now, that's a governance gap, not a bandwidth problem.
Read how it became institution-runWhat one person knows how to do hasn't been written down anywhere that survives them leaving the room.
If onboarding a new hire takes months of shadowing, or the same mistake gets made by a different person every time, the knowledge is still living in someone's head instead of a system.
See the documentation modelThe same initiative gets executed differently depending on who's running it this time.
If results vary by location or team even though the strategy on paper is identical, execution is running on individual judgment, not a shared standard.
Read the case studyReviews turn into information-gathering exercises instead of decision-making ones.
If your meetings run long, cover the same ground every time, and rarely end with a clear decision and an owner, the review process itself has become the bottleneck.
Read the case studyTwo engagements. Here's exactly what I built, and what changed.
Governed numbers still have to be read. That's the only place AI enters — after the system is already working.
On one of these systems, once the weekly numbers were already structured and reviewed, a layer was added that reads them and drafts a short brief before the meeting starts — what's on track, what needs attention, what to check first. It doesn't calculate anything; the same governed process still does that. It interprets what the numbers already show, and every recommendation gets checked against the same questions before anyone acts on it: is this actually supported by the data, does it have the full picture, does it match what's happening on the ground.
What changed is measurable in one place and not yet in another. Meetings now start with a structured read of the week instead of a search for one. Whether that shortens review time or improves decisions hasn't been formally tracked — so it isn't claimed here as more than what it is.
Every engagement follows the same five stages, in order — from first conversation to a system that doesn't need me in the room.
See the processI also write about the thinking behind this work — like why dashboards don't improve execution, and where AI should and shouldn't sit in a decision.
Read moreCurrently accepting a limited number of advisory engagements.
If you're building past the point your systems can hold you, let's talk about what needs to be built next.
Start a conversation