Team capability
Don't replace the team. Upgrade the way they work.
Generic AI training teaches people prompts they forget by Friday. This is built on the work your team actually does. We take real tasks from real roles, redesign how they get done, and put the systems, standards and confidence in place to keep it going.
AI should make good people better. Not replace everyone because somebody discovered ChatGPT.
This is usually how the conversation starts
“Everyone's using AI differently and quality is all over the place.”
“We paid for licences and half the team never logs in.”
“People are pasting client data into tools with no rules.”
“The team is nervous about AI and nobody says it out loud.”
“One person does everything clever and it doesn't spread.”
“We did AI training and nothing actually changed.”
You don't need to know the solution. That's the part I help with.
What we cover
Built on your roles, your tasks, your standards.
Role based upskilling
Sessions built around what each team actually does day to day, using their live work rather than generic examples.
Ways of working
The workflow redesign that makes AI worth using. New steps, new handoffs, agreed defaults for how the work gets done.
Prompt systems
Shared, reusable prompt libraries and templates tied to your brand, tone and process, so quality doesn't depend on who's typing.
Human review
Where a person must check, sign off or override. Protecting quality, brand, accuracy and the decisions that still need judgement.
Governance
Clear rules on data, confidentiality, client information, disclosure and acceptable use, written in language the team will follow.
Keeping capability
Internal champions, documentation and a review rhythm, so the knowledge stays with your people and not with an external consultant.
How it runs
Four stages, then it's yours.
Map the work
Understand what each role does, where time goes, and which tasks are genuinely worth changing.
Redesign the tasks
Rebuild those specific workflows with AI in the right place and human review at the right point.
Train on the real thing
Working sessions where the team applies it to live work, not a sandbox exercise they'll never repeat.
Make it stick
Standards, prompt libraries, champions and a follow up review to check what changed and fix what didn't.
What the team gets
The goal isn't fewer people. It's a more capable team.
- Redesigned workflows for the tasks that actually matter
- A shared prompt and template library tied to your standards
- Clear human review points and quality expectations
- Written governance on data, confidentiality and acceptable use
- Internal champions who can carry it after the sessions
- A follow up review measuring what changed in practice
Questions
The things people ask before we start.
Is this just ChatGPT training?
No. Tool training is the smallest part of it. The value is in redesigning the specific tasks your team does, agreeing where human judgement stays, and putting standards in place so quality holds.
How many people can take part?
Working sessions are most effective with small groups, typically six to twelve, so people can bring live work. Larger organisations run multiple cohorts by function.
What if our team is resistant to AI?
That's common and usually rational. Starting with their own frustrations, the admin and rework they hate, gets far better adoption than starting with the technology.
Do you handle data and confidentiality rules?
Yes. Governance is part of the work: what can and cannot go into a tool, how client information is handled, when AI use should be disclosed, and who signs off.
Can this follow on from the Diagnostic?
Often it does. The Diagnostic frequently shows that the biggest opportunity isn't new software, it's giving the team you already have better systems and the confidence to use them.