By the end of the course, students are able to distinguish the major families of AI (classical Machine Learning, generative AI, agentic AI) and to explain what each requires in terms of data, skills and human oversight. They can build and defend an AI-driven transformation proposal for a real organisation: defining an AI strategy, prioritising use cases, establishing a roadmap, allocating tasks between humans and AI, governance, legal compliance, an adoption plan and value measurement.
To enable students to understand the main types of AI and their applications, to define an AI strategy, to identify, experiment with and prioritise opportunities, to design and steer a transformation, and to integrate human, legal, ethical and environmental issues.
Seven thematic sessions covering the technologies (Machine Learning, LLMs, agents) and then the transformation itself (strategy, playgrounds and roadmap, portfolio and governance, responsible design, steering, law and adoption). Three sessions dedicated to a real-life company case worked on in groups, including two question-and-answer sessions with the “client”.
1. Classical machine learning
2. Generative AI: LLMs, prompting, RAG, hallucinations
3. Agentic AI: agents, orchestration, human-AI collaboration
4. AI vision & strategy: the “why” and the “how”
5. The “what”: Playgrounds, roadmap and portfolio
6. Responsible design: human-AI allocation, risks, bias, Green AI
7. Steering, governance, law and adoption: AI Act, GDPR, Intellectual Property, liability
8. Launch of the real-life cases
9. Client session 1
10. Client session 2
Short in-class exercises. Between sessions, individual preparatory work (for example, analysing an AI application announced in the press and identifying the underlying label), then, from session 8 onwards, group work on the case.
On-site classes alternating lectures, industry examples and short in-class exercises. From session 8 onwards, group work on a case inspired by an existing organisation, with access to the client through question-and-answer sessions: what students learn depends on what they ask.
Group presentation before a mock executive committee, followed by a defence. The proposal must cover the problem and its strategic importance, the solution and the human-AI allocation, feasibility, risks and compliance, environmental impact, the implementation and adoption plan, and the expected value and how it is measured. Students are expected to use AI tools, but such use must be declared and every statement must be justifiable without the tool.
Main material: slides from the ten sessions, distributed after each class. More specific references shared in class.
| Training | Study programme | Block | Credits | Mandatory |
|---|---|---|---|---|
| Master in Business Engineering | Finalité spécialisée en Sustainable & Digital Management | 1 | 3 | Yes |