Enterprise AI

There is a gap between technical teams and managers, in understanding and in misaligned incentives. We found this gap to be terribly costly, in money and in failed artificial-intelligence projects. Starting from that reality, we built this course from our practical and theoretical expertise.

To close this gap, we teach the fundamental concepts not only for your projects, but also for your strategic goals. You already agree that your company should use more AI. We spare you the biases and the traps! We work with one client company at a time, because it is in the precision of a single subject that you progress the most.

We love and know mathematics and programming, but we won't talk about that: there are so many other aspects that will let you succeed in your AI projects!

Three families of AI, three ways to create value

Three families of AI, three ways to create value

Machine Learning

Learn from experience rather than explicit instructions. Nobody explains how to ride a bike: you get on, you fall, you adjust.

Learn from experience rather than explicit instructions. Nobody explains how to ride a bike: you get on, you fall, you adjust.

Generative AI

Imitate creatively. We improvise like jazz, with structure. We produce and understand text, audio and images with highly convincing quality.

Imitate creatively. We improvise like jazz, with structure. We produce and understand text, audio and images with highly convincing quality.

Agentic AI

Multiply skills that interact with each other. A small prompt like “march on Berlin” and general Patton moves from North Africa across all of Europe.

Multiply skills that interact with each other. A small prompt like “march on Berlin” and general Patton moves from North Africa across all of Europe.

Multiply skills that interact with each other. A small prompt like “march on Berlin” and general Patton moves from North Africa across all of Europe.

A course built around AI in your company

A course built around AI in your company

Forget video calls: we work with a single client at a time, on site, whether that's a team of 2 or 20, so we can actually understand your situation and your needs.

Between the theory and the hands-on work, we leave room for real discussion. The workshops are built around you, our client. We're here to help you see what matters as a manager, even if you're technical yourself.

Get ready and bring your energy: the ship is leaving the dock and everyone's on deck. There's no room to doze off here, this isn't a long-winded lecture. It's about learning enough to get back to work and carry your projects forward. We'll make sure of it, sailor!

Jour 1, 2 et 3

To plan and tailor the content to your company's goals

Jour 4 et 5

Intensive deep dives: mornings devoted to theory, afternoons to hands-on workshops on your own challenges

Jour 6 et 7

Strategic wrap-up, with a tailored action plan for results that last

Who is this course for?

Whether you're a manager, a CxO or an AI specialist, this program helps you move past the buzzwords and bring AI into your company strategically, with a leadership style built around impact. Don't get bamboozled by the geeks. Don't let your boss frustrate you either. Success comes from communication and mindset between leadership and engineers, in both directions.

Let's talk about your needs

Let's talk about your needs

5+1

hands-on workshops + capstone

10

“Dangers” decks FR / EN

8

portraits of AI figures

36 h+

of course material

300+

slides, updated since 2022

The program, in detail

No technical prerequisites: we start from intuition and a real case, the concept comes after. Five parts, one workshop per part and a capstone.

(01)

Framing and the Compass

Where do you start and how do you avoid the wrong starting point?

  • Data is not truth. Why an exact figure can mislead and what that changes for decision-makers.
  • The Compass: perceive, decide, act. The full loop and the human's place at every notch.
  • The shift. From an AI that answers to an AI that acts: what changes in how work is organized.
  • The four eras of AI. A timeline to know where we stand, without jargon.
  • Context, not prompt. The bottleneck has moved: not writing speed, but the quality of framing and review.
  • The three Us: useful, usable, used. The sieve that separates a lasting project from a one-day demo.
  • Workshop: the usage map. Everyone places their own tasks on the Compass and spots where the human must stay.

(02)

Understanding AI without mathematics

What can this machine really do and where does it make things up?

  • How AI learns. The intuition of learning, without a single equation: examples, patterns, corrected mistakes.
  • The AI Palette. The families of tools and what each one can do, so uses are never confused again.
  • No data, no science. Why data quality and provenance decide the outcome, before the algorithm.
  • From smooth talker to grounded expert. How an AI is tied to verifiable sources so it stops improvising.
  • Hallucinations. Why a machine invents with confidence, how to spot it, how to protect yourself.
  • Workshop and deliverable. On concrete cases, tell what AI can do from what it pretends to know.

(03)

Deciding right

How do you measure the value and risk of a use case, in euros and minutes?

  • Measuring means first measuring how wrong you are. You don't measure success, you measure error.
  • From the confusion matrix to the business matrix. Translate every error into euros, for this specific business.
  • The asymmetry of errors. Not all mistakes cost the same.
  • The KPI a human understands. Pick the unit the business reads without a translator: a euro, a minute, an avoided defect.
  • Drucker versus Goodhart. What gets measured gets managed, but an optimized indicator stops being good. Hold both.
  • Perceived ROI and the iceberg effect. Why we think we are faster than we are and the hidden cost below the surface.
  • Two case studies: Klarna and Zillow. What a mistake at scale costs, documented and quantified.
  • Dialing autonomy. The crystal-clear rule: the more a mistake costs, the less autonomy the AI keeps.

(04)

Critical mind, critical awareness

How do you keep your judgement facing a machine that flatters and errs with confidence?

  • Describing is not deciding. The line a manager never crosses, even when the AI is convincing.
  • From critical mind to critical awareness. The critical mind judges the text in front of it; critical awareness asks who produced it and why.
  • The three diseases of AI. It forgets what was agreed, changes its mind between runs, promises without always delivering.
  • Resonating is not reasoning: sycophancy. Why nothing numbs judgement like a mirror that approves.
  • The taxonomy of six biases. Recognize the biases that slip into a use case, from hiring to recommendation.
  • Choosing your bias: the formal impossibility. You cannot optimize everything at once: a leadership decision, not a technician's.
  • The two extremes: Galactica and Gemini. Two opposite public failures, one shared lesson about measure and restraint.
  • Workshop. Spot, in your own exchanges with an AI, the moments it flatters, forgets or derails.

(05)

Governing AI that acts

When AI acts on its own, how do you keep control, safety and independence?

  • From answer to action. What changes when AI sends, books, orders, chains actions together.
  • The permissions dial and MAYA. Fine-tune what the AI is allowed to do and how far.
  • The fuse. Bound the action to bound the effect: committing the house, never without a human signing.
  • The five degrees of agency. A clear scale, from simple assistant to autonomous agent.
  • The risk formula: a product, not a sum. One neglected factor is enough to sweep everything away.
  • The seven rules of instituted agency. The frame that makes an acting AI governable.
  • The four defensive contracts. What you guarantee, what you forbid, what you trace.
  • Enterprise sovereignty. Host AI at home, depend on no single vendor, keep the trace. Sovereignty does not ask who owns the technology, but who depends on whom the day the wind turns.
  • The Ship of Theseus and the Pareto reserve. Lasting: keep live human competence even when the machine does the heavy lifting.

🎓 Capstone workshop: the business case before a board

Participants carry a real use case end to end: the business need, the value in euros, the autonomy dial, the guardrails, the indicator that proves. They defend it before a simulated executive committee. That is where the whole course comes together.

By the end of the course, a manager can…

Frame

Place an AI use case on the Compass and spot where the human must keep control.

Discern

Tell what an AI can do from what it invents, without mathematics.

Quantify

Translate a use case into value and risk: the cost of an error, in euros.

Dose

Set the autonomy dial: what to hand to the machine, what never to leave it.

Doubt

Recognize AI's flattery, its lapses, its biases — with method.

Guard

Set the guardrails of an acting AI: fuse, permissions, traceability.

Protect

Defend enterprise sovereignty and last over time.

Convince

Carry an AI business case end to end and defend it before leadership.

+14 years of AI expertise

Warith Resume

Warith Resume