Case Study: Fondation Aigues Vertes
Training the board and the management team before the rest of the institution
Situation
Fondation Aigues Vertes is a Geneva social and educational institution supporting people with disabilities. Its governing bodies wanted to form their own view of generative AI before letting it into the institution. Four obstacles kept coming up: fear of the tool combined with a lack of hands-on practice, no clear idea of what a paid version actually adds, difficulty seeing where AI would create a concrete benefit, and concern about its environmental footprint.
Objective
Give the board and the management team real practice rather than a lecture: how to choose a tool compatible with the processing of sensitive personal data, where AI genuinely changes the daily work of a social and educational institution, and what separates a connector, an automation and a true agent.
Overall satisfaction
98.5%
Net Promoter Score
+70.6
Promoters
82.4%
Format
4 half-days
Process
Four half-days, two groups
The board and the management team were trained over four half-days, split into two groups of twelve so that the sessions stayed interactive and everyone handled the tools rather than watching them.
Choosing tools when the data is sensitive
How to choose a working tool when the data involved is sensitive personal data about people with disabilities, the question of environmental footprint, and the concrete difference between MCP connectors, automations and true agents in production, illustrated with examples from the social and educational sector.
The use case built with them: the night warden assistant
A night warden has to record the events of his round in the in-house software, but by morning the details escape him. An assistant was built: he dictates his round during the shift, and the assistant writes it up in a format that can be entered straight into the software.
Results
Overall satisfaction of 98.5% and a Net Promoter Score of +70.6, across 17 respondents
82.4% promoters, meaning participants who gave a score of 9 or 10
A working night warden assistant: the round is dictated during the shift, the report comes out in the format the in-house software expects
Governing bodies now able to judge for themselves whether a tool suits the data it would handle
Testimonials
- Excellent in the energy brought to it and in the way the training was run, rich in examples, thoroughly practical. (Cynthia L.)
- Very dense training, almost too short. Great energy. Thank you very much. (Vincent G.)
- Participants differ in how much they know and practise, yet the training was still accessible to a beginner like me, thanks to the clarity of the explanations and of the materials. (Sonia G.)
- I was afraid this training would overlap with others I had taken before, but not at all. We went quite far and saw a lot of varied uses. (Fabien R.)
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