![Cover Image for Vector Search for E-Commerce: Predicting Customer Churn [Barcelona Meetup]](https://images.lumacdn.com/cdn-cgi/image/format=auto,fit=cover,dpr=2,background=white,quality=75,width=400,height=400/uploads/98/110859ea-ec77-4bbd-96be-9291bce98a7a.png)
Vector Search for E-Commerce: Predicting Customer Churn [Barcelona Meetup]
Context
Barcelona runs on marketplaces. Glovo, Wallapop, Adevinta, TravelPerk, Skyscanner, and Shalion all live or die by the same question: which customers are about to leave, and what should we show them before they do?
Most churn models look at a customer as a row of numbers. Vector search looks at them as a point in a space, next to every other customer who behaved the same way, and lets you ask what happened to those neighbors. Add a geo filter, a payload index, and filterable HNSW (Hierarchical Navigable Small World) and the question becomes: what happened to customers like this one, in this city, on this plan, in the last 30 days?
This meetup is about building that engine, and about what search looks like in a city that shops in two languages.
Program
We are joined by speakers from Qdrant and the Barcelona e-commerce community.
Our agenda:
5:30pm: Doors open | Food, drinks and networking
6:00pm: Welcome
6:10pm: A Customer Churn Prediction Engine for E-Commerce [Talk by Qdrant]
6:45pm: Multilingual and Multimodal Retrieval for Marketplaces: Spanish and Catalan listings, image-based product search [Speaker to be announced]
7:10pm: Community demos: three short demos from local builders
7:30pm: Breakout for networking
9:00pm: Doors close
What You'll Learn
How to turn behavioral events into vectors you can actually query. How constrained search (geo filters, payload indexes, filterable HNSW) keeps results relevant without emptying the result set or blowing your latency budget. How to score churn risk from nearest neighbors instead of training a separate model. And how to run one index across Spanish and Catalan listings, plus image queries, without maintaining three systems.
Who this is for
E-commerce and marketplace engineers, data scientists working on retention and personalization, search and relevance teams, and anyone in Barcelona building on vector databases. Talks are in English. No prior vector search experience required.
Coming from AI Summit Barcelona (September 22 to 23)? Stick around a week. This is the community follow-up.
Space is limited. Register early.