OpenSearch Project - München

Hi Everyone,
mark your calendars for the fourth OpenSearch Project München meetup

Agenda
6.30 pm Start (Drinks and snacks will be provided)

6.45 pm Aswath Srinivasan
Senior Search Engine Architect, AWS
# Budget Friendly Semantic Search With Neural Sparse Search
This Level 300 technical session is for those Search teams using BM25 and wanting to implement Semantic Search but has a budget constraints due to large vertical node requirements for implementing KNN search. This will be a short, straight-to-the-point talk with comprehensive reference material to get started quickly.

This talk will give a walkthrough on different best practices that should be considered/incorporated when implementing Neural Sparse Search in OpenSearch, starting with when to use sparse vectors vs. dense vectors. It will cover topics such as,

1/ determining when to use sparse embeddings vs. dense embeddings,
2/ pros & cons of hosting the sparse embedding models remotely & locally,
3/ pre-trained models available in OpenSearch and other possible open-source huggingface sparse embedding models,
4/ when to use document-only encoders vs. bi-encoders,
5/ Hybrid Search with Neural Sparse Search,
6/ GPU acceleration recommendations,
7/ some Neural Sparse Search benchmarks.
8/ Lastly, some “lessons learned“ so you don’t have to learn it the hard way.

The talk will include a 3 minute demo to set the stage for Neural Spare Search and will act as an Intro.

7.30 pm Break (Drinks and Snacks)

7.50 pm Niklas Bieda,
Developer, Freelancer - Eliatra

# Lexical and Vector Search with Open Search

Discover the ultimate search symphony where lexical precision meets semantic intelligence, delivering results that understand both what users say and what they mean. Unlock the future of search technology where these complementary approaches join forces to overcome limitations, creating experiences that are both powerfully accurate and intuitively relevant.

8.30 pm Networking

9.00 pm Close.