Skip to content
Module
Back to courses
AI & BusinessAll levels

Put AI to work in your business

A five-step method to delegate, describe, check and capture, applied to your real case.

Course designed by Alex Dionisio, Senior Product Manager. Updated September 28, 2026.

  • 54 steps
  • ~2 hr 10 min

Level: All levels · No prerequisites: accessible to everyone.

Choose which tasks to hand to AI, protect your data, write precise requests, check the results and turn one successful use into a team method. You leave with a first deliverable produced on your real case, a reusable workflow and your team's AI policy.

What you will be able to do

  1. Explain, on a concrete case, why an AI assistant makes things up, forgets or misses recent information, and decide how to use it accordingly.
  2. Sort data before giving it to an AI tool (allowed, anonymize first, forbidden) and configure the tool to match.
  3. Choose which tasks to hand to AI and how to work with it (automate, assist, delegate) based on value, risk and the data involved.
  4. Write a complete request (task, context, examples, format, constraints) and improve it over at least two documented iterations.
  5. Pick the right tool for a deliverable and justify the choice on data, cost, licensing and integration.
  6. Check AI output against an explicit checklist (facts, figures, sources, tone, compliance) before using it.
  7. Turn a successful use into a reusable workflow with human checkpoints, and set the team's rules of use in an AI policy.

Prerequisites

  • An account on a chat assistant (Claude, ChatGPT, Microsoft Copilot, Mistral Vibe or similar); a free plan is enough for the exercises
  • A real, recurring task from your own work in mind
  • No technical skills needed. To go deeper on prompting for product work: Learn prompting.
  • The legal references cited are French and EU (CNIL, GDPR, EU AI Act), checked on 28 September 2026; outside the EU, check your local equivalent (for example the ICO and UK GDPR in the UK). This course is not legal advice.

Syllabus

What will I produce in this course, and in what order?

  1. Objective · By the end of this introduction, you will know what you are going to produce (a deliverable on your real case, a reusable workflow and an AI policy) and how the three modules get you there.

Why does an AI assistant get things wrong, and what can I safely give it?

  1. Objective · By the end of this lesson, you will be able to explain why an AI assistant makes things up, forgets or misses recent information, and choose the right fix for each case.

  2. Objective · By the end of this lesson, you will be able to sort your business information into three levels (allowed, anonymize first, forbidden) and configure your AI tool to limit how your conversations are reused.

How do I get from a real need to a useful deliverable, in a way I can repeat?

  1. Objective · By the end of this lesson, you will be able to choose a task to hand to AI, decide how to work with it (automate, assist, delegate) and write the deliverable statement in one sentence.

  2. Objective · By the end of this lesson, you will be able to run the five steps of the method (frame, choose the tool, describe, check and iterate, capture and deploy) on a given case, and link each step to a competency of the 4D framework.

  3. Objective · By the end of this lesson, you will be able to write a complete request (task, context, examples, format, constraints) and explain what each element changes in the result.

  4. Objective · By the end of this lesson, you will be able to choose the right tool for a deliverable and justify the choice on four criteria: type of deliverable, data involved, cost and licensing, integration.

  5. Objective · By the end of this lesson, you will be able to check AI output against a five-point checklist and improve it with precise feedback, over at least two documented iterations.

  6. Objective · By the end of this workshop, you will have run the five steps on your own case and obtained a first checked deliverable, together with the request that produced it.

How do I turn a good result into a reusable, shared method without losing control?

  1. Objective · By the end of this lesson, you will be able to save a proven request as standing context (project, instructions, shared template) and document it as a workflow with its human checkpoints.

  2. Objective · By the end of this lesson, you will be able to choose between a page, a small app, a shared assistant and an automation for your need, and list what remains to be done before opening the result to customers.

  3. Objective · By the end of this lesson, you will be able to write a one-page AI policy (approved tools, forbidden data, permitted uses, checks, transparency, owner) and plan a gradual rollout in your team.

  4. Objective · By the end of this lesson, you will be able to plan your next 30 days of AI use (one workflow in production, a measured review, one extension) and find the course's reference resources.

Verified sources

The fast-moving facts of this course (tools, models, regulation) are checked against these sources, then reviewed according to how quickly they go stale.