talk-data.com talk-data.com

Event

Search Technology Meetup - Haystack edition

2025-09-22 – 2025-09-22 Meetup Visit website ↗

Activities tracked

2

We are very excited to announce another Search Tech Berlin Meetup. This time it will be a special edition on the evening before the Haystack conference. The meetup will be hosted and sponsored by GetYourGuide (thank you!).

Agenda

18:30 - Doors open 19:00 - Introduction 19:10 - Talk: Yasir Mohamed & Viktoriia Kucherenko: What we wish we knew when we started working on Search Suggestions 19:50 - Networking, food & drinks 20:10 - Talk: Trey Grainger: Beyond hybrid search: Creating wormhole vectors to bridge disjoint vector spaces 20:50 - Let’s socialize and grab some drinks 21:30 - Doors closing

Talks

1) What we wish we knew when we started working on Search Suggestions Yasir Mohamed & Viktoriia Kucherenko (GetYourGuide)

There’s a thousand and one ways to serve Suggestions in a Search bar, and in GYG we developed approach #1002. We’d like to take you through a brief journey of how our Search Suggestions evolved beyond basic Search-as-you-type-Suggestions in OpenSearch to effectively support Named Entity Recognition, Typo Tolerance and more.

Yasir's a Senior Software Engineer with GetYourGuide. Originally he joined the company to work on building the processing Infrastructure for the company's Search Engine, he quickly fell in love with the domain and has been working on improving Search ever since. His focus is on refining Search to match user behaviour; constantly iterating on GetYourGuide's search capabilities to always serve what users are looking for. Outside of work, he's an avid solo traveller, cook, and pint enthusiast.

Viktoriia is a Senior Software Engineer at GetYourGuide, where she has specialized in the Search domain for five years. She built an event-based system for search indexing that enables data flow from various teams across GetYourGuide, making it accessible through flexible Search APIs. Her expertise includes distributed systems, observability, and search technologies. Outside of work, she's a proud mom, Swiftie, and sports fan.

2) Beyond hybrid search: Creating wormhole vectors to bridge disjoint vector spaces Trey Grainger (Searchkernel)

Modern search systems increasingly rely on various vector representations—dense embeddings for semantic similarity, sparse vectors for lexical matching, and behavioral vectors from collaborative filtering. But these vector spaces typically operate in isolation, forcing us to choose between different notions of relevance.

This talk introduces the concept of "wormhole vectors"—a technique for dynamically connecting and traversing between heterogeneous vector spaces using retrieved documents as a bridge. Rather than running parallel searches and merging results (the most common hybrid search approach), we can start in one vector space (e.g., using sparse embeddings for keyword search), retrieve documents, and then derive wormhole vectors to instantly transport us to another space (e.g., using dense embeddings for semantic search) to continue searching with a fundamentally different notion of similarity.

We'll explore practical uses of this approach across three vector spaces: dense semantic embeddings, sparse lexical/keyword vectors, and behavioral vectors from matrix factorization. We’ll demonstrate how document sets can dynamically materialize bridges between these vector spaces, enabling novel retrieval patterns like semantic → behavioral → lexical traversal that surface user intent patterns invisible to single-space approaches.

This isn't just hybrid search—it's dynamic cross-space navigation that opens up new possibilities for query understanding, result diversification, and improved relevance.

Trey is lead author of the book AI-Powered Search (Manning 2025) and the founder of Searchkernel, a software consultancy building the next generation of AI-powered search. He also serves as a technical advisor at OpenSource Connections. He previously served as CTO of Presearch, a decentralized web search engine, and as Chief Algorithms Officer and SVP of Engineering at Lucidworks, an search company whose technology powers hundreds of the world’s leading organizations. Trey is also co-author of the book Solr in Action (Manning 2014), as well as over a dozen other publications including books, journals, and research papers. Trey has 18 years of experience in search and data science, including significant work developing semantic search, personalization and recommendation systems, and building self-learning search platforms leveraging content and behavior-based reflected intelligence.

See you at the event and at Haystack!

René & Roman

Sessions & talks

Showing 1–2 of 2 · Newest first

Search within this event →

Beyond hybrid search: Creating wormhole vectors to bridge disjoint vector spaces

2025-09-22
talk
Trey Grainger (Searchkernel)

Modern search systems increasingly rely on various vector representations—dense embeddings for semantic similarity, sparse vectors for lexical matching, and behavioral vectors from collaborative filtering. But these vector spaces typically operate in isolation, forcing us to choose between different notions of relevance. This talk introduces the concept of "wormhole vectors"—a technique for dynamically connecting and traversing between heterogeneous vector spaces using retrieved documents as a bridge. Rather than running parallel searches and merging results (the most common hybrid search approach), we can start in one vector space (e.g., using sparse embeddings for keyword search), retrieve documents, and then derive wormhole vectors to instantly transport us to another space (e.g., using dense embeddings for semantic search) to continue searching with a fundamentally different notion of similarity. We'll explore practical uses of this approach across three vector spaces: dense semantic embeddings, sparse lexical/keyword vectors, and behavioral vectors from matrix factorization. We’ll demonstrate how document sets can dynamically materialize bridges between these vector spaces, enabling novel retrieval patterns like semantic → behavioral → lexical traversal that surface user intent patterns invisible to single-space approaches.

What we wish we knew when we started working on Search Suggestions

2025-09-22
talk
Viktoriia Kucherenko (GetYourGuide) , Yasir Mohamed (GetYourGuide)

There’s a thousand and one ways to serve Suggestions in a Search bar, and in GYG we developed approach #1002. We’d like to take you through a brief journey of how our Search Suggestions evolved beyond basic Search-as-you-type-Suggestions in OpenSearch to effectively support Named Entity Recognition, Typo Tolerance and more.