KI-Experimente in Kriens

Last Thursday, 60 city employees packed an AI fair in Kriens to see seven experiments we built in nine months.

We started with a workshop. 14 ideas came out of it. The team voted. 7 made the cut. Each experiment followed the same process: Interview the idea giver, define the problem, build a prototype, test it.

A financial forecast planner: Worked. A chatbot for citizen services: Worked. Automatic meeting transcription: Didn't work. Two legal document research tools: Both worked. Recycling bin route optimization: Too complex. A web scraper detecting unregistered hotels and Airbnbs: Worked.

Here's what I'm still figuring out: how do you keep that momentum alive once the experiments end and the day-to-day takes over again?