PyData Berlin 2026 October Meetup

Welcome to the PyData Berlin October meetup! We would like to welcome you all starting from 18:00. There will be food and drinks. The talks begin around 18.30 and the doors will close at 18.30. Make sure to arrive on time! If you cannot attend, please cancel your spot so others are able to join as the space is limited. Host: Vinted is excited to welcome you to this month's version of PyData. We need you to enter your full name and email when registering for this event. You will receive a personal PIN that you need to enter on the reception tablet upon arrival.

Photos may be taken during the event. ************************************************************************** The Lineup for the evening: Talk 1: How Vinted taught search to understand what you actually meant Abstract: This talk will be about how we tackled a mismatch in search: what users want for a query and what the results page shows can drift apart. We implemented a small model that predicts what each query is really about (say 90% one type of item and 10% another). We then reorder results so the page reflects that mix, without touching the rest of the search system. In a Vinted A/B test, the items users were looking for gained visibility, engagement increased, and buyer satisfaction stayed healthy.

Speaker: Beatrice Musizza is a Data Scientist on the Search Relevance team at Vinted in Berlin, where she is the technical owner of the reranker. Her work focuses on large-scale reranking architectures, user modeling, and balancing marketplace discovery with user intent. She recently co-authored and presented research on query-level multinomial blending at the ACM RecSys 2026 OARS Workshop. Outside of work, she holds a judo black belt and is always happy to learn new languages!

Talk 2: The Surprising Power of Python yield: A Coordination Mechanism for Data Systems Abstract: Most Python developers know yield as the feature behind generators and lazy iteration. But the most interesting thing about yield is not that it produces values — it is that it suspends computation. In this talk, we'll start with familiar generators and progressively expand the model. We'll see how suspended computation can describe workflows rather than immediately execute them, how yield from naturally composes larger workflows from smaller ones, and how coordinating runtimes can pause, persist, and resume execution across failures and even across many short-lived Lambda invocations.

Speaker: Stefan is a Lead Software Engineer at a financial services provider. Previously, he spent eight years as an independent consultant, delivering data engineering projects across a range of industries. Earlier in his career, he worked at several Berlin-based startups and Microsoft. He holds a PhD in Information Retrieval from Northeastern University in Boston.

Lightning talks There will be slots for 2-3 Lightning Talks (3-5 Minutes for each) between the two main talks. Kindly let us know if you would like to present something:) *** NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery are not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. If you haven't yet, please read the detailed version here: https://numfocus.org/code-of-conduct ***