AI workshop for your team, on your own tasks

How the days run

• We start from the tasks, not the features. For each role, three to five tasks that recur in their week, written in their language — “turn the client call notes into a follow-up email before I leave for the day”, not “summarize text”.
• We design backwards from the task. One objective per task, one worked example on realistic data, one practice run on the team’s own material, and a written definition of what a good output looks like in their context.
• We work on your real material. A flashy demo that leaves people unable to repeat it on their own files teaches the tool, not the job.
• We finish with the test that matters. Can a person in that role complete one named task the day after, without me in the room?

One day is enough for one team and a handful of tasks. Two or three days let several roles work in parallel, or add the layer between being trained and being competent — the reinforcement that keeps a task in use after the session.

You leave with

• Three to five recurring tasks per role, written down as tasks.
• A worked example and a practice run completed on your own work.
• A written definition of a good output per task, so people can judge their own.
• Whatever was configured or built during the session, kept for your team.
• A short plan for what to reinforce afterwards, so the tasks stay in use.

Format