AI Leadership Training is a fixed-fee session, half a day or two half-days, for owners, COOs, CFOs and heads of function. It runs on your own processes and numbers. They leave with a shared way to cost a process, a way to price a vendor's claim, one page of ground rules, and a checklist for reviewing anything AI produces.
Who this is for
Every department head has an AI vendor in their inbox and nobody on the leadership team can tell a good pitch from a bad one.
You have said yes or no to AI spend on instinct and you would like a method you can repeat.
Your board expects the leadership team to have a view on AI and the view so far is a feeling.
Something AI-produced has already gone out of the building unchecked and you do not want that again.
What you get
A list you can check you received.
- A shared costing method the whole leadership team uses the same way: volume, minutes, loaded cost, delay.
- A way to price a vendor's claim: what to ask for, what a real answer looks like, when to walk away.
- One page of ground rules for AI use in your business, drafted in the room and owned by someone named.
- A review checklist for anything AI produces before it leaves the building, from a customer email to a report.
- Two or three of your own processes costed during the session, which double as worked examples.
- The pre-read and exercises, kept, so a new head of function can be brought to the same level later.
How it runs
- 01
Pre-read: we ask for two or three of your processes and any AI proposals currently on the table.
One week before
- 02
Session one: costing your processes together, then pricing a real vendor claim from your inbox.
Half a day
- 03
Session two, if taken: drafting the ground rules and the review checklist on your own examples.
Half a day, one to two weeks later
- 04
Written pack: the method, the rules, the checklist and the worked examples from the room.
Within one week
What we will not do in this programme
This is training in judgement, not in tools. Nobody leaves knowing how to write a prompt, deliberately. If you want your team to operate AI software, this is the wrong programme.
Who this is for, and who it is not
This is for the owner, the COO or CFO, and the heads of function. Sales, operations, finance, service. It is not for your engineers or your IT lead, and it is not for the people who will operate the software. Those people need something different, and most of it they can get from the vendor.
The leadership team’s problem is not operating AI. It is deciding about it. Every head of function has a pitch in their inbox. Each pitch has a number in it. Nobody in the room has a shared way of testing that number, so the decision comes down to who argues best, or to instinct, or to whichever vendor’s reference customer sounded most like you.
Why judgement, not skills
We train judgement because skills date and judgement does not. The software your team might learn this year will be different next year. The method for costing a process, pricing a claim and deciding what needs a person will be the same. We also train judgement because it is what protects you from us. A leadership team that can cost its own processes will know when our proposal does not add up, and we would rather work with that team than with one that takes us on trust.
The first half-day
You send us two or three of your own processes a week ahead, and any AI proposals currently on the table. The first half-day opens with costing those processes together, on a whiteboard, using the working we use in every engagement: volume, minutes per item, loaded hourly cost, downstream delay. Your heads of function will disagree about the inputs. That disagreement is the point, because it is the first time the room has argued about a number rather than a feeling.
Then we take a real vendor claim from your inbox and price it. What is the claim actually saying? What would have to be true for it to hold? What would you ask for to check it, and what does a real answer look like? By the end, the room has a shared vocabulary and one worked example it produced itself.
The second half-day
If you take the second session, one to two weeks later, it turns the vocabulary into rules. The team drafts one page of ground rules for AI use in your business: what data may go where, which decisions still need a person, who approves a new use. Then it drafts a review checklist for anything AI produces before it reaches a customer, a regulator or the board. Both are written in the room, on your examples, and both leave with a named owner.
What usually goes wrong with AI training
Most AI training for leaders is a keynote. It is interesting, it is general, and a month later nobody can point to a decision it changed. The opposite failure is a hands-on workshop that teaches executives to type into a chat window, which is a skill they will not need and will forget.
We avoid both by running the session on your numbers and your inbox. Nothing in the room is hypothetical. The processes are yours, the vendor pitch is one you actually received, and the rules are ones you will actually have to follow on Monday.
What your team can do afterwards
Cost a process the same way, so two heads of function bring comparable numbers to the same meeting. Read a vendor’s proposal and know which question to ask first. Say no to an AI use that breaks the ground rules, and say yes quickly to one that does not. Judge our own proposals with the same method, which is why the title of this page says what it says. The judging itself stays with your people. That was the purpose.
What usually comes next
Most teams take the ground rules straight into AI Governance and Controls, or use the costing method to decide whether the AI Strategy Roadmap is worth commissioning.
Next in the sequence: Custom AI AgentsWhere this programme applies
The one ask on this page
Ninety seconds now. Thirty minutes if it is worth it.
Take the check, see your score with no email asked for, and decide whether the thirty-minute review is worth your time. If you already know the process, skip the check and tell us.