On Tuesday 9 December, PyData Eindhoven was once again all about data, machine learning and practical stories. With 23 talks, a broad community and a special PySport track, it was an inspiring day for data and AI professionals.
We are particularly proud that two Pippelaars took to the stage themselves this year.
From global retail planning to detailed football analysis
Yannick Mariman and Harlley Lima (IKEA) presented how retail planning is organised on a large scale with Metaflow. In their talk, they showed how sales forecasting, fulfilment planning and capacity assessment, each with their own models and teams, are cleverly linked together. By orchestrating workflows and making them reproducible, more than thirty countries can collaborate effectively within a single coherent planning chain. The shared patterns and solutions are recognisable to any organisation with complex, interdependent ML processes.
Annemarijn Blom then took the audience to the football pitch. In her presentation, she showed how tracking and event data can be used to automatically recognise playing styles in the crucial moments before a shot on goal. By modelling players as spatial networks and applying techniques from network science, attack and defence patterns become objectively transparent, without time-consuming manual video analysis. This opens up new possibilities for analysts and coaches to understand and improve tactics and playing behaviour.
Learn, share, and contribute to the community
In addition to the two talks, a total of ten Pippelaars were present at PyData Eindhoven: as visitors, volunteers and speakers. It is precisely this combination that makes events like this so valuable: sharing knowledge, gaining new insights and working together to build the data community in the region.
Thanks to PyData Eindhoven for organising the event and to all visitors for the good conversations. On to the next event.
