Our app recommends you items to put on your to-buy Migros list. The suggestions are based on the intelligent analysis of your shopping history (i.e. Migros Cumulus data), while our models are learned from the data, and can thus incorporate changing consumer behavior, out-times (e.g. holidays) and uncertainty (e.g. through global models based on shopping histories of the other consumers).
We programmed a prototype front end for Android smartphones as well as a prototype back end on Google Cloud Platform. The software was tested on real data as well as on plausible generated data.
Register as alpha tester: https://groups.google.com/forum/#!forum/vergissminoed
Hack Zurich 2014
https://www.hackerleague.org/hackathons/hackzurich-2014/hacks/vergiss-dot-mi-dot-noed