AI Training Programs for Employees
Built on the work your people already do: real tasks, real files and the tools your firm already allows. Up to ten people per session.
A course on the tool is not a change in how the work gets done.
Most AI training teaches features in the abstract: what the tool can do, a few prompts, a demo. Then people go back to their desk, open the same files and do the job the way they did it last week.
A program works when it starts from the other end. Pick the work your team actually does, decide which parts AI should take, which it should help with and which stay human, then practice that on real files until it becomes the normal way of working.
Five things to look for in any AI training for employees.
Real tasks
One workflow your team runs every week, not a generic exercise.
Your approved tools
Copilot, ChatGPT or Claude, whichever your firm allows, with the files people are allowed to use.
A way to sort tasks
A shared language for what AI does, what you do together and what stays human.
Checks before output is used
An agreed review checklist, so AI output is checked by someone who knows the work.
Something to repeat
Worked examples and a guide, so the workflow survives after the trainer leaves.
Four ways of working with AI. Every task goes in one of them.
In the training, people sort their own tasks into SCAN. It is how a team decides what to hand over and what to keep.
Substitute
Hand it to AI
Turn notes that are already checked into your standard report layout.
Complement
Do it together
List the questions a client file still leaves open, then decide which to ask.
Aid
AI helps you think
Brush up on an unfamiliar rule or product before you advise on it.
Non-negotiable
Stays with a person
Talk a worried client through a difficult decision.
Train your team privately, or start with a public session.
One workflow, built for your team
Your team picks a single workflow and up to ten people take part. You supply sample files you are allowed to share and access to your AI tools; the exercises are built on them before the day.
Several days, building on each person's workflow
For teams that want more than one session. A specialist consultancy ran a three-day Copilot program for ten people across research, recruitment and finance.
Live workshop: how to use AI without losing your brain
For independent consultants, AI transformation advisers and fractional AI leads. Work through SCAN on three tasks from your own week. Dates and syllabus on Maven.
Recorded webinar: Thinking with AI
An introduction to SCAN and to deciding when to hand over a task, work with AI or do the thinking yourself. Watch the recording.
After each private session your team keeps worked examples, written steps for running the workflow again and a checklist for checking what AI produces.
Example: finding what's missing before a client meeting.
The job
Before a client review, an adviser has to know which details and documents are still missing. The session used the firm's real client form and its own checklist for meetings.
What AI did
Copilot checked the form against the checklist, listed the missing items and out-of-date documents, and quoted the text behind each one. One of them was the client agreement.
What the adviser did
A missing entry is not always a missing document. The adviser confirmed whether it was filed somewhere else, whether this review needed it and what had to be updated.
What clients say
"It's amazing how much better you guys think about AI than we do."
Start with one workflow and up to ten people.
Say which work you want to change, and the session is built around it.