Case Study: Union Cycliste Internationale
From the executive committee to the teams: structuring AI adoption end to end
Situation
The Union Cycliste Internationale, the international cycling federation based in Aigle in the canton of Vaud, combines constraints rarely found together: several working languages, teams spread across very different functions, an events calendar that leaves few windows, and data part of which concerns athletes' private lives. Generative AI was arriving desk by desk, with no decision taken as a whole.
Objective
Take back control of an adoption that was spreading from the bottom up: have the leadership decide rather than unmanaged usage, measure the real state of practice before training anyone, then build the team training on those findings rather than on a generic syllabus.
Steps in the engagement
4
Starting point
The executive committee
Tool selected
Claude
Interim deliverable
Audit report
Process
Leadership first
The executive committee was trained before the teams, so that it could make decisions on concrete grounds rather than on what it had read in the press.
Audit, then audit report
A map of what already existed: tools in use, real practices, expectations and risk areas. The written report then served as the basis for decisions and as the starting point for the training programme.
Level 1 training on Claude
Fundamentals of generative AI, its limits and its main business use cases, advanced prompt engineering, then an in-depth module on the Claude ecosystem: Web and Desktop versions, Projects, Artifacts, Skills, memory, Claude Code and Cowork for multi-step tasks, plus integrations with Office, Chrome, Monday, Teams, Outlook and SharePoint. The session closes with advanced use cases, an innovation challenge applied to the UCI, a quiz and a data protection segment.
Results
A full sequence run in the right order: leadership trained first, the state of practice measured next, team training built on the audit findings
An audit report documenting real usage and serving as the basis for management decisions
Teams trained on a single tool, Claude, and on its integrations with the software already in place rather than on one more browser tab
A data protection segment built into the training itself rather than treated as a legal appendix
An innovation challenge applied to the UCI, turning the session into production rather than a lecture
Contact us
- Want to upskill your teams?
Let's schedule a meeting to assess your training needs
- Want to identify high-ROI use cases for generative AI?
Let's schedule a strategic call