LEARN / WORKFLOW PREPARATION

Bring one workflow into focus.

Before choosing an AI tool, describe a task that happens repeatedly. A short account of real work is more useful than a long list of features.

1. Follow a real example

Write down what starts the task, what arrives, who handles it and what a completed result looks like. List each tool touched along the way.

Illustrative example: an enquiry arrives → someone checks for missing details → a colleague prepares a quote → a person approves the reply.

2. Find the friction

Where do you enter the same information twice?
What do you wait for, chase or check manually?
Which mistakes create rework?
What changes on a difficult or unusual day?

3. Pick one observable measure

Use something you can record before and after a change: time spent handling each enquiry, the number of follow-up rounds, or incomplete briefs. Record workload and unusual cases too, so a quiet week does not masquerade as an improvement.

4. Decide what must stay with a person

Identify decisions with financial, customer or information-handling consequences. State what should happen when information is missing or a tool gives an uncertain answer.

5. Bring a safe example

Use a fictional or redacted example for an initial discussion. Remove customer names, contact details, account numbers and confidential content. Keep passwords and private records in your own systems.

Your short brief: “When ___ happens, ___ needs to do ___. The difficult part is ___. We would judge an improvement by ___.”

This guide is general workflow preparation, not a promise that a particular task should be automated.

Email Matthew