From Coupons to Clicks: Recommender Systems at Migros & Digitec Galaxus

The Digitec Galaxus Product Development Meet-Up is back — this time diving into recommender systems with two talks bridging data science and engineering practice.

Talk 1: Insights from Building and Operating a Recommender System for Migros Supermarkets
Max Nolte – Data Scientist & Team Lead at Supermarkt IT, Migros-Genossenschafts-Bund

When you think of a Migros supermarket, you don't usually think about recommender systems. Over the past four years, my team and I have built and operated the engine behind personalized Cumulus coupons and a few other digital touchpoints. I'll share how it works, why we invest in it at a brick‑and‑mortar retailer, and what we've learned, covering data, models, measurement, and the practical realities of running it day to day.

Talk 2: Recommendations at Digitec Galaxus — From Data Science to Engineering Practice
Michael Weiss, Senior Software Engineer & Robert Rosenbach, Senior Data Scientist – Digitec Galaxus

How do you go from a promising model to a recommender system that is fast, scalable, and maintainable in production? We will walk through the Digitec Galaxus recommendation stack end to end: the engineering architecture that keeps recommendations snappy and reliable, and the data-science side of model training — what to optimize for, which input signals matter, and the trade-offs involved in both.

The two presentations will take roughly a combined hour, leaving plenty of time for discussions, questions, and exchange afterwards over drinks and snacks. Whether you come from a software engineering, data science, product, or any other background — we'd love to have you join us!