Search Technology Meetup XVII: Restart!

Let's restart our Meetup again! After a long break we will start directly with three great talks by spending a great evening together.
A lot of technologies, features and things changed in the past, but our focus will stay the same: Search Technologies.
Thanks to Tudock GmbH for the drinks and the location.

Doors are open at 6:30 pm (including food & drinks). We will start at 7pm.
Sounds like a great event to you so far?
Then you also need to check out these great talks (all will be in english):

Querqy query rewriting for vector search
Many search practitioners are currently wondering whether they should try out vector search in their applications and how they could set up an experiment for it quickly. At OpenSource Connections we have recently added vector search to Chorus - a template e-commerce search application based on open source libraries. The idea is to make it easy for search developers to get started with their experiments on vector search quickly.
In this talk, I will briefly introduce vector search in Chorus and then focus on how vector-based queries are added to the search query by help of the Querqy query rewriting library. An understanding of Querqy’s query rewriting will be key for developers when they want to progress and add their own modifications to the Chorus template setup. I will show how a query string is turned into a vector that can then be used for the retrieval of search results or for boosting in a hybrid search approach that combines vectors with keywords. Furthermore, I will discuss how the Querqy query rewriting framework will allow for combining vector search with manually maintained rules, such as for boosting and filtering results - a prerequisite for applying vector search in e-commerce applications.

Speaker:
René Kriegler works as Director E-commerce at OpenSource Connections. He looks back at 16 years of work experience in search, including projects for some of the top 10 German e-commerce sites. René is co-founder and co-organiser of MICES (Mix-Camp E-commerce Search), an event that brings together the e-commerce search community each year. His technological focus is on Solr, Elasticsearch and Lucene. He created and maintains the Querqy open source library for query rewriting. René is co-initiator of the Chorus project – an open source software stack that combines Querqy with other powerful tools to build e-commerce search and to measure and improve search quality.

Decoupling query rewriting and learning to select from search engines
With increasing maturity regarding query preprocessing and learning to select algorithms in a company, needs are growing for distributing the development of features as well as to decouple them from the search engine itself. This is why we developed Querqy-Unplugged, an enhancement of the Querqy ecosystem that applies query rewriting outside of a search engine in order to decouple the processing and the deployment of rules as well as learning to select features from the indexing and querying load in the search engine. Querqy-Unplugged furthermore builds queries in a way to enable using vanilla releases of search engines without the need of custom deployments including plugins, and therefore additionally aims to facilitate the hosting part of maintaining search engines.
Querqy-Unplugged is maintained together with Otto, where it is included in a query preprocessing service that is running on top of the search engine. The values of this decoupling are faster and easier deployments, facilitated testing as well as a faster integration of additional learning to select algorithms. Furthermore, this architecture enables the prospective foundation of a dedicated query preprocessing team.
The talk will discuss the advantages of the new architecture, give insights about challenges regarding the migration as well as describe several learning to select algorithms handled by the query preprocessing service.
Speaker: Johannes Peter & Volker Carlguth

How we cut in half our AWS search budget using Java Flight Recorder
E-commerce-search engines are resource eating monsters. Fast response times are key thus a lot of both heap and cpu is needed. For a customer we were challenged with reducing their overall AWS budget spent on driving their Solr cluster. In this project report we'll navigate through a classic cloud architecture way down to the foundations of garbage collection in the JVM. We'll explain the benefits of using ARM over Intel processors. We achieved a breakthrough by analysing the Solr internals using the Java Flight Recorder, upgrading the JDK to version 19 and experimenting with the Z garbage collector. In the end it worked: we cut the customers's AWS bill in half.

Speaker: Dennis Berger & Torsten Bøgh Köster