A hands-on AI adoption workshop for BNF Group, run inside the company's strategy retreat at the Ritz-Carlton Singapore on Mon 30 Mar 2026. 40+ senior managers, ages 40 to 55, almost none of whom had ever opened an AI tool. 3 hrs, taught on Claude, laptops out.
Published 9 September 2026
40+ senior managers from BNF Group, ages 40 to 55, in a function room at the Ritz-Carlton for the company's strategy retreat. Operators rather than technologists, each of them running a real part of a real business.
Almost none of them had used an AI tool. Not "tried it once and let it go". Never opened one.
A strategy retreat with 3 hrs carved out of it, and a room of senior managers who had never used an AI tool. That was the brief, and it sets the bar at AI 101 whether you like it or not.
So no vibe coding, no agent architecture, nothing that needed a subscription bought first. We taught on Claude and participants worked on free accounts, which meant the only thing standing between them and using it on Tuesday was knowing what to type.
Bringing in someone from outside helps with a room like this, and it isn't about the material. Experienced managers will push back on an internal recommendation and then sit politely through the same point from a stranger. I'd rather have the pushback, so I go looking for it early.
3 hrs on a Monday afternoon, laptops open from the first hands-on block.
The plain-language version, with live demos rather than slides about slides. Then a way to choose between the tools instead of collecting all of them, and a proper look at Claude, since that is what they'd be using on Tuesday.
One analogy carries this whole block for a room of managers: prompting is briefing a new hire on their first day. They already know how to do that well. They just haven't tried doing it in a chat box.
Groups take a real piece of work from their own side of the business and put it through Claude. Not a sample dataset, not a case study. The thing that was already sitting in their inbox that week.
Each group builds one thing they can use again, a Claude Project or a skill for their own function, and presents it to the room. Peer learning does more here than another demo from me would.
Groups worked on paperwork from their own part of the business, which means no two runs of this session produce the same thing. Ordinary documents, the kind that sit in an inbox all week and never quite get written. That is the point. A generic exercise teaches the tool, and a document you actually owe someone teaches the habit.
The demos ran on the same principle. Claude took a spreadsheet of mock sales data shaped like theirs and turned it into something you could read at a glance, then pulled 3 separate reports into one synthesis while the room watched. Senior managers don't need to be told AI is powerful. They need to watch it do the admin that eats their week.
By the end of the afternoon a room of 40+ senior managers who had never opened an AI tool that morning were building on their own work and presenting it to each other. That is the whole result, and I'm going to leave it there.
No survey went out after this one, so there are no confidence scores or satisfaction numbers on this page. There is a measured version of the same format on the Claude Masterclass write-up if you want figures before you book.
One statistic in my draft deck did not survive the fact-check and came out. Every number I put in front of a room now has to trace back to a source first, and so does every number on this website.
