Skip to main content

Enterprise AI

The programme for steering your AI projects, as a two-day bootcamp or a full course, on your premises.

Deciding with AI, fully informed.

Every AI makes mistakes. This programme teaches your teams to put them to work: priced in euros, the error becomes the tool that steers an AI project, from the first framing all the way to a satisfied client.

Thirty minutes to talk about your project, then a written proposal.

A river in a wooded gorge opening onto pale hills, a tiny rowing boat under an amber sky

They trusted me. The course has been taught in the MBA of Rennes School of Business since 2022.

Sound familiar?

Executive leadership

We have launched several AI projects. Which ones actually pay off?

Your teams price what each project brings in and what its errors cost. You decide on an amount, not on a demo.

HR and learning

Everyone uses AI, nobody knows what they are allowed to do with it.

A frame role by role, the works council consulted at the right time, adoption measured team by team.

Transformation and innovation

Our proofs of concept never make it to production.

A threshold signed by the business, a measured autonomy for each agent, a file that convinces the board.

Every AI makes mistakes. Our method uses them as a project management tool.

(1)

Price the error

With your own figures, we set in euros what each error costs and what each right answer brings in.

(2)

Tune the machine

That price sets the threshold: what you hand to the AI, what a human checks, what you keep for yourself. The same reasoning tells how far an agent may act on its own.

(3)

Steer and defend

The project is judged by what it brings in, where the client feels it. Everyone leaves with a method they apply on their own, and a project they can defend before their board.

You are…

What you get

  • Managers who price in euros what an AI brings in and what its errors cost, then turn that into a threshold that steers the project.

  • A clear rule for every agent: what it is allowed to do, who signs off, how a human takes back control.

  • A deployment file and a pitch per participant, built on your real projects, handed over to you at the end.

Six formats, tailored to your sector

Talk

Length
1 or 2 hours
For
executive board, offsite
Your teams leave able to
look at AI in your sector with one precise question

Choose among the 42 talks ↗

Deciding with AI

Length
half a day
For
executive board
Your teams leave able to
say why AI this year, then price the error that steers the project

Steering AI

Length
one day
For
business unit heads
Your teams leave able to
price the errors, then set how far an AI may act on its own

The bootcamp

Length
two days
For
managers
Your teams leave able to
assess a project, write an agent’s spec sheet, build their deployment file

The common core

Length
four days
For
managers, project leads
Your teams leave able to
the whole path, from the first model to the pitch before the board

Getting AI adopted

Length
one day
For
HR, transformation
Your teams leave able to
decide, job by job, what to hand to AI, consult the works council, then measure adoption team by team

Every format is tailored to your sector (health, finance, industry and energy, public sector, media, e-commerce) and to your cases. In-house, on your premises, one client at a time.

Why this programme is different

  • The figure that costs, the steering wheel of the project. Accuracy invites debate; the price of an error invites a decision. The threshold follows from the ratio of two costs you sign off on, and the project is judged by what it brings in.
  • Permission to act. An AI that acts is governed by what it is allowed to do, not by what it is politely asked.
  • Evidence before opinion. Every figure in the course states whether it is measured, estimated or reported.
  • Open source, the backup generator. The day a vendor cuts you off, you keep running on your own machines.
  • A reservoir tailored to you. More than one hundred and fifty hours of polished course material, in French and English, to meet your needs.

Who teaches

Warith Harchaoui, PhD in applied mathematics (generative AI), ENS Cachan MVA. Interim head of AI for CAC 40 and SBF 120 clients, consultant at :probabl. (the scikit-learn company), teacher of the “Enterprise AI” course in the MBA of Rennes School of Business since 2022.

Every trade is a language: I learn yours, I share mine.

Warith Harchaoui speaking at a lectern, microphone in hand

What clients and students say

Group photo at the end of a session, in front of a screen reading THANKS

Clients

“Warith is a bible in AI”

Nicolas Renard, AI Manager, Jellysmack

“With Warith, glasspop makes ophthalmology history”

Dr Florent Costantini, Ophthalmologist, CEO of glasspop

“I used to use ChatGPT for my AI questions, now I ask Warith. There is no stopping progress”

Dr Alexandre Larmagnac, Head of Pre-Engineering, Volpi Group

“Warith's expertise in AI and Computer Vision has been foundational for our R&D team”

Maxime Schacht, CEO of VizioSense

“This cooperation with university thanks to Warith brought great value to the company I founded, Oscaro, leading to the Cerbero product”

Pierre-Noël Luiggi, Founder of Oscaro.com, Oscaro Power and xBrain

“Thank you Warith for your excellent job!”

Nathalie Birocheau, Founder of Ircam Amplify

Students

“AI for Business, a must! Skills shared with passion. Thank you Warith! Ready to take action!”

Charlotte Hérault, EMBA 2024, Deputy Head of Entity, Société Générale

“Thank you Warith for your enthusiasm and for sharing your thoughts and knowledge. We learned a lot in class over these two days. Fabulous course!”

Aurélie Amelot, EMBA 2025, Audiologist Manager, Audilab

“This collaboration with you is like fine wine: it improves year after year because it is made with excellent ingredients and by excellent professional for excellent students!”

Irena Descubes, Senior Manager of International Executive Programs, Rennes School of Business

“It was a huge pleasure!!! Thank you, Warith, for this wonderful experience.”

Philippe Marie, EMBA 2025, Business Unit Manager, BOCCARD Food & Ingredient

“Thank you Warith for the inspiring course during the core modules of our eMBA program!”

Jordy Brouwer von Gonzenbach, EMBA 2024, Independent Project Manager

“What an exciting time! Thank you Warith for these inspired and inspiring first steps in AI.”

Roger Le Dreau, EMBA 2023, Founder of Sea Val Solutions & Mai Waters

How it works

Dr Florent Costantini, CEO of glasspop, and Warith Harchaoui side by side behind their laptops during a working session

(1)

A thirty-minute meeting

Your sector, your projects, who will be in the room.

(2)

A written proposal

The format, the length, the sector and the chosen cases, with the quote.

(3)

The session, on your premises

On your cases, with one client at a time. Each participant's deliverables go to the sponsor.

What you receive

The slides
as a PDF, at the end of the programme.
A participant workbook
the cost matrix, the agent sheet, the deployment file, the pitch.
Each participant’s deliverables
handed over to the sponsor.
A certificate
or a grade, for a school.

The first step: let's talk

Frequently asked questions

Do I need a technical background for the Enterprise AI programme?

No. Deraison’s Enterprise AI programme requires no code and no compulsory formula. It is for executive boards, business unit heads, managers, project leads and HR teams; every idea arrives through a case from your sector.

Our teams already use ChatGPT: what does the Enterprise AI programme add?

Your teams already have the tool. The Deraison Enterprise AI programme teaches what a tool does not say: what each of its errors costs in your business, what you can hand to it, what a human checks, and how to judge a project by what it brings in.

Do we need an AI project already?

No. Without a project, the Enterprise AI programme starts from your processes and draws out the cases worth pursuing; with a project, it prices it and puts it to the test in the room.

How many participants does an Enterprise AI session take?

A Deraison Enterprise AI session brings together 5 to 20 participants. Below five it is an interview with slides; above twenty it is a concert.

Where does the Enterprise AI programme take place?

Deraison’s Enterprise AI programme takes place on your premises, in-house, with one client at a time. The case you work on is one of your own.

Does our data leave the room?

No. During Deraison’s Enterprise AI programme your cases are worked on in the room and your data stays there: the exercises run on your own machines, on open-source code, with no subscription to an online service.

How is the programme content set?

The Enterprise AI programme is set in two steps: a first thirty-minute meeting, then a written proposal that states the format, the length, the sector and the cases chosen.

Another question?