How to Run an AI Workshop for Your Team (Agenda Included)

A good AI workshop is 90 minutes, hands-on, and built around tasks your team actually does. The aim is for everyone to leave with one working prompt or workflow they will use the next day, not a slide deck they forget. Below is an agenda you can run as-is. The difference between a workshop that changes how people work and one they forget by Friday is almost always the same: real practice on real tasks instead of a lecture about possibilities. Keep the talking short, get people doing the work early, and capture what they produce in a shared library so the value outlives the session. The plan that follows covers preparation, a minute-by-minute agenda, and the follow-up that makes the skills stick.

Prep before the room

• Pick two or three real tasks the team does often.
• Confirm everyone has access to an approved tool.
• Gather a few real examples to work from.
• Set one clear goal: each person leaves with a usable prompt.

The 90-minute agenda

1. 1

   Frame the why (10 min)

   Explain the goal and the guardrails. No theory dumps..

2. 2

   Live demo (15 min)

   Show one task done with AI, thinking out loud as you prompt..

3. 3

   Guided practice (30 min)

   Everyone does the same task on their own example..

4. 4

   Share and improve (20 min)

   Compare results and refine prompts together..

5. 5

   Save the wins (10 min)

   Add the best prompts to a shared library..

6. 6

   Next steps (5 min)

   Agree on what each person will try this week..

Keep it concrete with real work

Workshops fail when they use fake examples. Use a real customer email, a real report, a real quote. The Google Cloud use-case library is a useful source of realistic scenarios to adapt to your industry.

Make the skills stick afterward

One workshop is a start, not a finish. Schedule a 20-minute follow-up in two weeks to share results and unblock people. Our guide to training a skeptical team covers the cultural side that keeps adoption alive.

Common mistakes to avoid

• Too much talking, not enough doing.
• Fake examples that do not match real work.
• No follow-up, so the habit fades.
• Skipping guardrails, so people build unsafe habits.

Turn the workshop into a habit

A single session is a spark, not a fire. What keeps the skills alive is the routine you set up afterward: a shared prompt library people contribute to, a short weekly check-in for the first month, and a champion who answers questions as they come up. Without that follow-through, most of what people learned in the room fades within a couple of weeks, and you are back where you started. Decide on the follow-up before the workshop ends, while motivation is high.

Set one clear, modest expectation for the weeks after: each person tries the new approach on at least one real task and reports back. Broad workforce research such as the WEF Future of Jobs report points to rising demand for practical AI skills, so the habit you build here is worth protecting. If some attendees remain wary, our guide on training a skeptical team covers how to bring them along without pressure.

Run it again as people and tools change

One workshop covers the people in the room on the day, but teams gain new members and tools gain new features. A short repeat session every few months, or a lighter monthly share where people show one trick that helped, keeps skills current without constant retraining. New hires especially benefit from a quick version during onboarding, so they start with the same habits and prompt library as everyone else rather than improvising their own.

Keep these follow-up sessions even shorter and more focused than the first. The goal is not to re-teach the basics but to spread what is working and surface what is not. Let the team drive the agenda by asking what they have tried and where they are stuck, and the sessions stay useful instead of turning into a chore people skip.