Massive Scale OpenSearch + Vector Search for AI @ GenAI Loft TLV
Details
Join us for an evening of real-world OpenSearch stories, hard-won lessons, and practical insights — straight from the teams pushing it to the limit.
📍 Location: Amazon Experience Center, Floor28
🕕 Time: Doors open 18:30 — talks start 19:00
***
### Agenda
🤖 Vector Search Made Simple: Getting Started with OpenSearch for AI Applications
- Dotan Horovits, AWS
OpenSearch has evolved from search to a powerful vector database solution that supports trillions of vectors and tens of thousands of QPS. This talk will begin by explaining the journey from lexical search to semantic search, and how OpenSearch combines both approaches in one complete package.
The session introduces fundamental concepts of vector databases, including how they store and process embeddings of various data modalities—whether text, images, audio or video—and use exact and approximate k-NN algorithms to calculate proximity.
We'll explore practical applications such as visual search, semantic search, and recommendation engines, with emphasis on real-world use cases. This is your opportunity to learn how OpenSearch can serve as a knowledge base for AI systems, particularly in applications like retrieval augmented generation (RAG) with large language models.
🧨 Oops, We Hit a Petabyte: How Shield Broke and Rebuilt OpenSearch the Right Way
- Omer Eckstein, Shield
You don’t reach petabyte scale without a few explosions along the way. This is the story of how Shield’s engineering team took OpenSearch beyond its comfort zone - and lived to tell about it. From disastrous indexing strategies to late-night cluster resurrections, we’ll explore the biggest mistakes we made and the principles that emerged from them. You’ll leave with honest insights, a few laughs, and a set of proven best practices for scaling OpenSearch without losing sleep (or data).
####
***
Come for the tech talks, stay for the networking, pizza, and good vibes with fellow engineers, data folks, and search enthusiasts!