[Masterclass for Pro] Modern generative AI Skills for Scientists: From Everyday Use to Custom Workflows

Overview

Generative AI tools are transforming how research and science is done, written up, and taught. But using them well appropriately requires more than casual prompting. This training gives scientists, educators, and students a structured, hands-on path from understanding the current generative AI ecosystem to building custom, automated workflows that support real research and teaching work.

Across two days, participants will learn to move beyond simple chatbot interactions toward using AI as a reliable, critical co-thinker: a tool that can be taught your own skills, connected to your documents and systems, and trusted to carry out complete workflows while remaining alert to its limitations, including hallucination, data protection, and reproducibility concerns. The spirit of the training is practical and critical rather than promotional: participants are expected to evaluate every AI-generated result rather than accept it at face value.

Target Audience

This training is aimed at researchers, faculty, doctoral and master’s students, and technical/administrative staff in the biomedical sciences who want to integrate generative AI meaningfully into their scientific, educational, or writing workflows. No prior programming experience is required for Day 1; Day 2 involves more technical, developer-oriented tools and assumes comfort with the concepts introduced on Day 1.

Structure: Two Days, Day 2 Optional

The training is organised into two sequential days:

Hands-On, Vendor-Aware Approach

Both days involve practical, hands-on work: participants will actively use generative AI tools during the sessions rather than only observing demonstrations. Bring a laptop and be ready to install and configure software before and during the training.

Throughout the training, concepts are illustrated primarily using Anthropic Claude and GitHub Copilot, as these currently offer the most complete tooling for the workflows demonstrated. A paid subscription (e.g., Claude Pro/Max, GitHub Copilot) is recommended for the best hands-on experience but is not mandatory ,  participation is possible with free-tier access, though some exercises may be limited without it.

Importantly, this is not a single-vendor training. While Claude and Copilot serve as the working examples, the instructor will also discuss equivalent tools and approaches from other providers (such as OpenAI ChatGPT, Google Gemini, and Mistral), so that participants can transfer the concepts to whichever ecosystem they ultimately choose to work in.

Registration

Billetterie Weezevent

Registration for this training is by application, not automatic enrolment. Places are limited, and admission is subject to approval based on:

  • Total number of available spots

  • Priority given to students and personnel of the biomedical sciences department (UFR des Sciences Fondamentales et Biomédicales)

  • A brief statement of motivation submitted with your application

Applying does not guarantee a place; applicants will be notified of the outcome ahead of the training dates.

  • Apply for Day 1 (registration link — Part I)

  • Apply for Day 2 (registration link — Part II; requires prior participation in Day 1)

Practical Information

  • Dates and times   Day 1: November 13th 2026 9:30-18:00 and Day 2 : November 16th 2026 9:30-18:00

  • Location Campus Saint Germain des Prés, Université Paris Cité, 45 rue des Saint-Pères, 75006 Paris

  • Application deadline November 6th

  • Maximum number of participants  24

  • Language of instruction  English

  • Cost  free for Université Paris Cité members; 500€/day for external participants

  • Contact  anthony.raffoul-rahme@u-paris.fr

  • A Certificate of participation from the Biomedical Engineering graduate School will be delivered.