Neos Conference 2026Boost your search by embracing Vector Search

Our journey from traditional string search to a semantic vector search using MariaDB. You'll see how product search improved and the impact of different, even local, embedding models and performance.

Our e-commerce platform had a relative basic string-based search with room for improvement. That's where we started to experiment with vector-based search using vector-databases, different remote and local embedding models. This boosted the search and made it much more robust. Will provide code samples, used infrastructure and a small guide on how to reproduce.