Workshop · Spring 2026 · Eastern Switzerland

    KI in der Hauswirtschaft

    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

    Block 1 · Opening

    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.

    Block 2 · Idea wall

    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.

    Block 3 · Practice

    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.

    Block 4 · Open questions

    Honest Q&A

    From 'What data am I even allowed to enter?' to 'How do I convince my supervisor?' — everything on the table.

    Block 5 · Take-aways

    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

    From the deck
    From the deck
    From the deck
    From the deck
    From the deck
    From the deck

    Tools used

    ChatGPTMentimeter

    The outcome

    50
    Participants reached
    5
    Use cases demonstrated
    2.5h
    Workshop duration
    100%
    Left with their own prompt

    "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

    Cookies & privacy

    We use analytics to keep making schnellstart.ai better. You decide what's on.