Choose a task before choosing a tool
List a few routine activities: turning meeting notes into action lists, drafting a first version of an internal update, or organising ideas for a presentation. Pick a task with a repeatable process and a result someone can check. Agree who will review it and what information may be used in the exercise.
For a mixed team, record each person’s starting experience. A beginner may need time to understand the interface and evaluate a response; an experienced user may need help connecting several steps into a reliable workflow. Ask how the course accommodates both.
Five questions to ask a training provider
- What will participants actually do? Ask for an example exercise and the time allocated to hands-on practice.
- How will they evaluate the output? Look for checks against the source material, completeness and suitability for the intended reader.
- What access is required? Confirm the laptop, account and subscription requirements before the session. Product access and features can change.
- How does the programme address workplace use? Discuss your organisation’s approved tools, information-handling rules and review process with the trainer.
- What happens afterwards? Ask participants to repeat one relevant workflow at work and record the time and review effort it needs.
Sample exercise: notes into an action register
The following fictional notes provide a small task with a checkable answer. They are a practice brief, not a client project or a reported training result.
“The team agreed to test a shorter weekly update. Mei will draft the agenda by Friday. Arun will ask two departments for feedback; no deadline was agreed. The first trial date is still to be confirmed.”
Ask your chosen AI tool to organise the notes into columns for action, owner, deadline and unresolved question. Tell it to mark missing information as “not stated” and to use only the supplied notes.
Then check the output yourself. It should preserve Mei’s Friday deadline, leave Arun’s deadline unspecified, and keep the trial date unresolved. A neat table is not enough if it invents dates or turns an open question into a commitment.
Measure the complete workflow
Try the task manually and then with AI. Record drafting time, checking time and corrections. For a fair comparison, use similar examples and the same quality checklist. Treat one successful attempt as a starting point and repeat the exercise before drawing conclusions.
Useful follow-up questions include whether colleagues can repeat the process, whether the reviewer can trace statements to the notes, and whether the result is ready for its intended use after review. These checks help distinguish an appealing demonstration from a practical workflow.
Download the fictional notes and review checklist (.txt)
Match the programme to the intended work
Welingtons’ one-day AI Productivity Playbook covers prompting, content, data and business workflows. For people specifically interested in building applications, compare the two-day Vibe Coding 101 course. Read Yavinesh Rajagopal’s trainer profile and share your team’s tasks when requesting a quotation.