AI in Home Economics
How 50 housekeeping leaders learned in 2.5 hours to use ChatGPT for staff scheduling, multilingual communication, and admin relief, FADP-compliant, practical, ready to use from Monday.
The starting point
The OdA Eastern Switzerland spring conference brought together around 50 housekeeping leaders, vocational trainers, and operations managers from care homes, hospitals, and catering businesses. The brief: make AI tangible, no tech-speak, no empty promises.
Participants arrived with the four questions we hear in every industry: Will I become redundant? Am I even allowed to use this? Where does my data go? And honestly, does this actually help in daily work?
What we did
Reading the room
Who has tried ChatGPT, who never has? From 'What is AI anyway?' to 'You decide what the AI sees, not the other way around' in ten minutes.
Six use cases from real shifts
From emergency staff scheduling at 6 a.m. to a multilingual notice for a team speaking five languages. Real examples from care homes and hospitals, not from a textbook.
Live on the projector and on every phone
GCES prompting, data-protection check, three examples every participant could follow on their own device. Quiet practice, not hands-up performance.
Honest Q&A
From 'What data am I even allowed to enter?' to 'How do I convince my supervisor?', everything on the table.
Five prompts for your desk
Every participant left with a one-pager and five tested prompts for their own operation.
Methodology
The GCES method
Goal · Context · Examples · Style, four building blocks that turn 'Write me a notice' into a concrete, usable answer.
Data-protection framework
One rule per data category: what never goes into the chat, what goes in pseudonymised, what's safe. FADP-compliant and practical.
Four-fears model
We address the four fears that show up in every beginner audience, job loss, permission, data, real value, before we show any tools.
What participants took home
Staff scheduling
Shift plans and emergency notices in minutes instead of hours, with templates that adapt to care-home and hospital realities.
Multilingual communication
A2-level notices in five languages for teams from Portugal, Eritrea, Sri Lanka, and Switzerland, without a translation agency.
Administrative relief
Email drafts, meeting minutes, SOP templates, everything that usually happens between two doorways.
Data protection in practice
Clear rules on what's allowed and what isn't. Not a theoretical FADP lecture but a checklist for the next prompt.
Your own practical prompts
Every participant leaves with five tested prompts for their own operation.
From the deck
Tools used
The outcome
"We took the room from 'What is AI?' to 'I have a prompt I can use tomorrow' in 2.5 hours. That was the goal, and that was the standard."
Lukas Huber, schnellstart.ai