Your leaders have had the AI briefing. They still do not know what to do on Monday.
Two days, sixteen hours, for leaders and team leads rather than executives only. Everyone brings one workflow they actually own, and it travels with them through all four modules: mapped, measured, skill-checked, then turned into something they try the week after. They walk out with four finished artifacts about their own work instead of a folder of slides about the future of it.
What it is not
Worth saying first, because most of what gets sold under this title is one of these three.
- Not a tools demo. Tools change every quarter. We work on the account and the approval boundary, which is what actually decides whether anything you build survives contact with your own organisation.
- Not a motivation session. Nobody gets told to embrace anything. Resistance from someone with no spare capacity is a rational response to being handed more work, and the playbook treats it that way.
- No promise that nobody's job is at risk. I cannot make that promise and neither can your managers, and a leader who makes it anyway spends credibility they will need later. We rehearse saying the true version instead.
The four modules
Day one is modules one and two. Day two is three and four. One workflow runs through all of them, which is the only reason the four artifacts add up to anything.
The AI-First Leadership Mindset
Where machine execution, human review and human authority each belong in a workflow you own. Everyone maps one, on paper, with the governance fields filled in rather than waved at.
They leave with: A completed workflow canvas, plus a better version of one real output they produced themselves, with a stated quality check and the review time counted.
The Landscape, and What Is Actually Evidenced
How to tell an independent trial from a vendor case study, and how to locate your own team honestly rather than against a maturity chart somebody invented.
They leave with: A stage diagnosis with the specific constraint that sets it, and an impact scorecard for the workflow from module one, measured by something other than a usage count.
Leading Skills and Role Transformation
Audit one role against what the work is becoming. Separate what the role will require from what you have actually observed, and tell a training gap apart from a permission gap.
They leave with: A completed skills audit for one role, finished in the room, plus a plan to test it with the person who does the work. No names, no scores about people.
The Change and Adoption Playbook
Diagnose resistance by asking instead of labelling. Sequence a change, place the human checkpoints deliberately, and rehearse the conversation you have been putting off.
They leave with: One honest conversation rehearsed against an observation rubric, a rewritten rollout message you could still stand behind in six months, and a first adoption action with a stop criterion.
What a participant walks out with
Not a certificate. Five specific things, each checkable by someone who was not in the room.
- Improved and tested one workflow they own.
- Defined an impact measure for it that is not a usage count.
- Identified one supported skill-development step, and classified the gap correctly.
- Rehearsed one honest conversation, including the part where the objection turns out to be right.
- Chosen one owned first action, with a criterion for stopping it.
Who it is for
Leaders and team leads who already have people, a backlog, and someone on the team who has asked them straight out whether this affects their job. It works best in a mixed room: engineering, data, security, risk, operations and the support functions together, seated by function so the examples land. It is not an executive briefing, and it is not an introduction to prompting.
If you cannot write a stop criterion, you have not defined success precisely enough to detect failure.
From the impact scorecard, day oneWhy mine and not the one with better slides
Because every external number I use is on a public page with its sample size, its setting and what is wrong with it. While I was building this, the most quoted study in my own material was superseded by the people who ran it. I rewrote three documents and published the correction rather than keep the better-sounding number.
That is the actual skill the two days are teaching. Your leaders are going to be sold AI claims for the rest of their careers, by vendors and by their own enthusiasm. Being able to take one apart is worth more than any framework I could hand them.
Running it for your team
It was built for a corporate client and runs as two cohorts, for leaders across engineering, data, security, risk and operations. It needs a room where the tables move, a confirmed answer on which tools your people can actually log into, and twelve to sixteen participants. More than that and the conversation practice stops working.
Tell me what your leaders are stuck on and I will tell you whether this is the right thing. If I am not the right person, I will say so on the call and point you at who is.
Aleodor Tabarcea · Site Lead at Stripe Romania. I build fast, autonomous teams by treating trust as an efficiency metric. Verifying every step causes latency; fearing mistakes causes packet loss. More · Advisory
Keep on learning, keep on building.🧡