A focused pilot.
A useful decision.

1. Choose the use case

Start with a department, a recurring task and a specific learning need. Identify the person responsible for the pilot and the people who will take part.

• One team and one defined use case
• A decision-maker and an internal point of contact

2. Agree on the scope

Before starting, agree on duration, participants, content, support and cost. Establish the baseline and the evidence you will review.

• Initial assessment and learning objectives
• Measures for participation, practical work and feedback
• A review date and decision criteria

3. Learn and practise

Participants follow learning activities connected to the selected task, with opportunities to practise and receive feedback. The exact learning content is defined in the pilot proposal.

4. Review and decide

Compare the agreed evidence with the starting point. Discuss what worked, what needs improvement and whether a wider rollout is justified. A pilot does not create an automatic subscription.

A concrete example

A support team could trial AI-assisted FAQ replies using fictional enquiries and approved public information. The review might compare accuracy and total time, including human review, and gather feedback on what support colleagues need. These are proposed measures, not promised results.