Prerequisites
No technical skills are required.
Upcoming Intakes
1 and 2 October: face-to-face sessions in Paris Villejuif & 22 and 23 October: synchronous online sessions
Price
€1,450.00 excluding VAT
Location
Paris–Villejuif
Professionnel
11
Links
Presentation
Target Audience
Objectives
Programme
Trainers
Isabelle de Padua
Executive Education Manager

Background

What if you could delegate your time-consuming tasks to an AI assistant capable of browsing the web, managing your files, drafting your emails and carrying out dozens of actions in a matter of minutes, whilst you focus on what really matters?

Claude Cowork, launched by Anthropic in 2026, represents a paradigm shift in the way we work. But to make full use of Cowork, you need to understand what’s happening ‘under the bonnet’: how does an AI agent plan? Why does it sometimes fail? How can you give it instructions that really work? This course answers these questions, moving from real-world examples to underlying concepts, and then back to advanced practical application. This course gives you the keys to make the most of it straight away.
IA agentique & Claude Cowork

Assessment and certification

To validate the training, each participant submits a personalised agent-based AI roadmap: a concrete plan for integrating AI agents into their professional practice, at the level of their role or organisation.
This deliverable provides an opportunity to consolidate what has been learnt and to leave with a clear, actionable vision tailored to their specific context.
Info

Find out more about the programme’s assessment and teaching methods  

See more

Who is this course aimed at?

No technical skills required.
For all professionals who want to take back control of their time. No technical skills required; if you know how to use a computer, you can take this course.
  • Managers and executives who want to improve their decision-making efficiency
  • Consultants and project managers faced with ever-increasing volumes of information
  • HR, marketing and support staff who handle repetitive, high-value-added tasks
If you manage information, draft documents, coordinate projects or carry out market research, you’ve come to exactly the right place.
Equipment required: One laptop per participant (Mac or Windows), an internet connection, and access to Claude Pro.

Why now?

Agent-based AI is no longer just a promise — it’s an operational reality. Organisations that train their teams today will gain a head start that will be difficult to catch up with. In 12 months’ time, the question will no longer be “should we adopt AI agents?” but “how have you trained your staff?”
  • A 2-day face-to-face course, 100% practical, where you’ll work on your own real-life cases, not fictional exercises
  • Suitable for all levels, with tailored teaching methods

Objectives

By the end of the training, participants will be able to:
  • Explain how an AI agent works (perceive–plan–act–observe loop)
  • Use Claude Cowork independently for their business use cases
  • Build multi-step automation workflows with Cowork
  • Diagnose and correct unexpected agent behaviour
  • Evaluate and choose between various agent-based tools (Cowork, Operator, Copilot, open-source alternatives)
  • Identify the risks associated with using AI agents in a professional context
Day 1 – From discovery to foundations
Module 1
Module 1
Cowork in action

A live demonstration of Claude Cowork covering various use cases. In 90 minutes, you’ll see the bot carrying out tasks that used to take you half a day. It really grabs your attention and sparks a desire to learn.

Module 2
Module 2
Guided first steps

You’re getting to grips with Cowork for the first time, step by step. First automation successfully completed first thing in the morning: research, summarising, formatting. The focus is on confidence, not on comprehensiveness.

Module 3
Module 3
Understanding agent-based AI

The conceptual core of the training: how an agent plans, makes decisions and acts. You’ll understand what goes on ‘under the bonnet’ — so you can navigate Cowork with precision and anticipate its limitations, without jargon or maths.

Module 4
Module 4
The art of agent-based prompting

Giving an agent clear instructions is a specific skill. You’ll learn about the TASK framework and rewrite your own prompts during the workshop so that they can be put into practice straight away.

Day 2 – Control and extension
Module 1
Module 5
Advanced workflows with Cowork

The whole morning is devoted to intensive practical work. You’ll build real-world workflows: information management, communication, reporting and multi-application workflows. Each workshop is based on your own business context.

Module 6
Module 6
An overview of agent-based tools

Cowork, Operator, Copilot, open-source alternatives
 You can compare the tools available on the market and draw up your own decision-making matrix to choose the right solution for your specific situation.

Module 7
Module 7
Risks, ethics and governance

Having an agent act on your behalf raises serious concerns. You can identify the specific risks (data, GDPR compliance, cascading errors) and work together to draw up a responsible use policy for your organisation.

Module 8
Module 8
Action plan & conclusion

The training doesn’t end in the classroom. You’ll leave with 3 workflows to implement from the following week, a pack of ready-to-use resources, and a 4-week post-training follow-up programme to help embed these practices over the long term.

Trainers

  • Dr. Tawhid CHTIOUI
    Info

    Areas of expertise
    Management, Management Accounting, Communication and Organisational Change, Informal Control and Panoptic Control, Governance

    Dr Tawhid CHTIOUI

    Professor of Management | Founding President and Dean of Aivancity
  • StĂ©phanie RODRIGUES
    Info

    Areas of expertise
    Artificial Intelligence, Generative AI and Agent-Based AI

    Stéphanie RODRIGUES

    Artificial Intelligence Trainer
  • Dr. Antoun YAACOUB
    Info

    Areas of expertise
    Generative AI, large-scale language models (LLMs), machine learning and deep learning, language and emotion analysis...

    Dr Antoun YAACOUB

    Head of the MSc in Generative Artificial Intelligence