Claude training for your team

It rhymes with them on purpose. My four-competency spine — delegation, description, discernment, diligence — references the open AI Fluency framework Anthropic published with Prof. Rick Dakan and Prof. Joseph Feller, and I will happily point your team at Anthropic Academy for the self-paced material because it is genuinely good. The curriculum and exercises on the day are mine. What a free course cannot do is run on your documents, in your room, with your team's actual objections in the air, and refuse to move on until it lands — self-paced completion rates are famously low and adoption rarely follows. That gap is the whole job.

Often you should not — if your work lives in Outlook, Teams and SharePoint, Copilot is the better first move and I will tell you so. Claude earns its place for long-document reasoning, drafting quality, and agentic work through Cowork and Claude Code. Plenty of organisations run both deliberately. I train on all four major platforms and have no reseller relationship with any of them, so the recommendation is honest.

On Claude Team and Enterprise plans your conversations are not used to train the models, and Enterprise adds SSO, audit logging and data-retention controls. The real risk in practice is not the vendor, it is your own people pasting the wrong thing into the wrong tool. That is a policy and habit problem, and we cover it on the day.

MCP is the open standard that lets Claude talk to your other systems — CRM, file storage, analytics, ticketing — with permissions you control per connector. Most teams do not need it on day one. Teams that want AI doing real work rather than answering questions need it by about month three, which is why it is on the curriculum rather than in a footnote.

Both, and the building came first. Archivers.ai and PlanningConstraints.com are live products built Claude-first, this website was built with Claude Code, and my own commercial operation — CRM work, reporting, research — runs through Claude with MCP connectors into the systems I use daily. When someone asks what breaks at scale, I am answering from last week, not from the documentation.

A workshop day works best at 8–14 people; beyond that hands-on time suffers and it becomes a lecture. Executive briefings work at any size because the format is different. If you need to cover more people, the sensible pattern is one briefing for leadership plus repeated workshop days by department.

On-site anywhere in the UK — I am based in Derby, so the Midlands and the North are easy days, and London, the South and Scotland are all regular. Remote delivery works well for briefings and follow-up sessions; for a hands-on workshop day, in the room is materially better.

Everyone leaves with three committed use cases and a 30-day plan, and you get the prompt and project templates we built on the day. Most clients then take a block of embedded support — office hours, agent building and adoption chasing — because the gap between a good day and changed behaviour is follow-up, not enthusiasm.

No — I am not a levy provider or a registered training body, and I would rather say so plainly than send you round the houses. If funded training is the priority, an apprenticeship route through a registered provider is a genuinely good option and will cost you less. What I do is a different purchase: one senior specialist, on site, changing how your team works within a day or two, invoiced as a normal consultancy engagement. People come to me when they need the capability this quarter rather than next year.