
How I think about security in my AI operating system
An authority ladder, the difference between a guardrail and a boundary, and the fake email my agent plants in my inbox every week to see if my assistant falls for it.
I build the human and AI operating systems purpose-driven organizations run on, and I teach leaders and teams to build this capacity for themselves. I got here through two decades inside organizations, most recently as an embedded fractional COO.
For organizations, I work with leadership teams on the systems they run on and on how their people use AI, starting either with the leader or with two sessions where the whole team writes its own agreement on AI. For individuals, I teach TEMPO, a six-week training cohort where you build your own AI operating system. No technical background needed.
Already know what you need? Book a free 15-minute call.
Most people already use AI as a thought partner. They open a chat, talk through a problem, and get something useful back. But the work still gets done the way it always was. Inside teams I see the same thing on a bigger scale: nearly everyone experimenting on their own, one or two people racing ahead, and no shared agreement on what AI is for or what it's allowed to touch.
I've built AI into how I actually work, as a colleague woven through everything I do. It's raised the quality, the range, and the amount of what I can produce, well beyond what I could reach on my own.
The research now shows both directions: people who accept whatever AI hands them get worse at thinking. People who direct it get better. Every system I build keeps the decisions with you. That is what I help people and organizations do.
Your organization
I work with leadership teams at purpose-driven organizations, building the systems their team runs on and teaching everyone to use AI well. It can start with the leader or a workshop for the whole team, then grow into a team program, a build for one process, or six months working inside your team.
You
I teach non-technical professionals to build their own AI operating systems, in a six-week cohort or one to one. You start from a system I've spent months building and refining, then make it yours.
In September I started Foundations with two teams, each working out where AI helps their work and where its limits are. One of them, a global health organization, finished both sessions with a written agreement on how they'll use AI, drawn from the team's own words.
Read how it went →As Tandem's fractional COO, I built a system for the team's campaigns: AI drafts the creative, an agent grades every piece, and the team approves at every gate.
Read the case study →I moved a fast-scaling nonprofit's contractor management, procurement, invoicing, and grants out of email and spreadsheets into one system that its staff and contractors work in.
Read the case study →The TEMPO operating system I built to run my own practice. Every client, project, and priority in one place, wired to the tools I work in and updated on its own. See the case studies →
Four case studies: an operations portal, a marketing automation system, my own AI operating system, and a whole team writing its own rules for using AI. I built the three systems without an engineering team.

An authority ladder, the difference between a guardrail and a boundary, and the fake email my agent plants in my inbox every week to see if my assistant falls for it.

People worry AI erodes critical thinking. Two days after I asked Claude one of my favorite questions, a colleague and I had a dashboard that answered every budget question we'd been stuck on.

It saved us almost half the train fare and booked our Airbnb. Then it asked for my children's passports, and its privacy policy didn't back up what it told me.
TEMPO is my six-week training cohort where non-technical professionals build a working AI operating system shaped around how they actually work. You work in a small group, with a live session every week (all of them recorded) and your own work running through the system from the first week.
You start from a system I've spent months building and refining, and you make it yours.
Next cohort: Tuesdays, Oct 13 to Nov 17
See TEMPO dates and joinEverything on this page comes down to one idea: adopt AI deliberately, so it strengthens your thinking, in service of work that matters. That's what I mean by AI on purpose.