The short answer: A regional leadership team spent one afternoon setting data guardrails first, then each leader built a working AI teammate on their own real tasks rather than a demo. Every attendee left with a saved Project, a live data connection, and a tested prompt, and every attendee rated the session 10 out of 10. The same working-session format is available as a team workshop.
The situation
A regional organization brought its entire leadership team together for a hands-on AI working session - what their team later called an "AI Afternoon." It is the pattern I see everywhere right now: the tools arrive before the training, and a leadership team needs a shared answer to the two questions that matter - what is this actually good for in our work, and what should never be pasted into it.
The brief was simple: no theory, no slideware. Everyone leaves with something they will still be using the following week.
What we did, in order
1. Data guardrails first
Before anyone typed a prompt, we set the rules. I use a green/yellow/red framework: green data is safe to use with AI tools, yellow needs care or stripping of specifics, red never goes in. Doing this first matters for a leadership team in particular - they are the ones who set the norms everyone else will follow, and "we covered safety before we covered tricks" is the difference between an AI policy people follow and one they route around.
2. Everyone builds their own AI teammate
The core of the afternoon: each leader built a real, reusable setup - not a demo. Each person left with:
- A Project with their standing instructions - the context about their role, their standards, and their preferences that the AI should always know, written once so every future conversation starts smart.
- A live data connection - the AI wired to information they actually work with, not toy examples.
- A reusable prompt, tested on real work - built, run, and refined in the room against a task from their actual week, so they saw it produce something useful before they left.
That last part is the whole design philosophy: the test of a workshop is not what people nod along to, it is what still gets used seven days later. A prompt that already worked on your real work on Tuesday gets used again on Monday.
The result
Every attendee rated the session 10/10. One review from the team, verbatim:
"Best afternoon of practical learning and application. Highly recommend and will be using him again. We brought our entire leadership team together for an AI Afternoon!"
That review sits with the others on my 5.0-star Google profile, alongside the 200+ professionals I have trained.
Why this format works for leadership teams
Most AI training fails leaders in one of two ways: it is either an executive briefing (interesting, nothing usable) or a tools tutorial (usable, but pitched below the decisions leaders actually make). The working-session format threads that needle because leaders do not just learn to use AI - they leave with a shared vocabulary for governing it. When the whole leadership team has personally built with the tools and personally applied the data guardrails, AI policy conversations stop being abstract.
Want this for your team?
This format is a team workshop - remote or in-person, scoped to your team size and tools, a fixed price set on a call. If you are not sure a workshop is the right first step, start with an AI Workflow Assessment or just book a free 20-minute call and tell me what your team is working on.
Frequently asked questions
What is an AI Afternoon?
It is a hands-on working session where a leadership team sets data-handling guardrails first, then each person builds a real, reusable AI setup on their own actual work rather than watching a generic demo.
Why start with data guardrails before covering any AI tools or tricks?
A leadership team sets the norms everyone else in the company will follow. Covering what is safe to use with AI, and what should never be pasted into it, before covering any tricks is what makes an AI policy people actually follow instead of route around.
What does each leader walk away with after the session?
A Project containing their own standing instructions, a live connection to data they actually work with instead of toy examples, and a reusable prompt that was built and tested against a real task from their own week.
How is this different from a typical AI training for executives?
Most AI training for leaders fails by being either an executive briefing that is interesting but not usable, or a tools tutorial pitched below the decisions leaders actually make. This working-session format is hands-on like a tutorial but pitched at real decisions, so leaders leave having built something rather than just watched a demo.
Did the leadership team actually rate the session well?
Every attendee at the session covered here rated it 10 out of 10, and the review sits alongside the others on Stephen Thorn’s 5.0-star Google profile.