Presented by:

Img 5398

Gleb Otochkin

Google

Gleb is a Cloud Advocate at Google who specializes in Cloud Database technologies in the Google Cloud. He has over 20 years of experience in data-related technologies, and has expertise in a variety of areas, including relational databases, Big Data, application development, data replication and integration solutions, including products from Google, Oracle, AWS, Cloudera, and other vendors. He has been honored to be a presenter at various conferences in the USA, Europe, and APAC.

Profile pic

Alan Li

Google

Alan Li leads the semantic search effort across GCP Database products. Previously, Alan worked on Spanner database internals and led efforts to scale transactional limits and silent data corruption protection. And prior to Google, he spent much of 15+ years building ML serving infrastructure or building databases.

No video of the event yet, sorry!

The Vector data type is one of the most used data types for the latest AI applications prototypes and implementations. It is commonly used in applications such as image retrieval, natural language processing, and recommender systems. Vector search is a technique used to find similar vectors in a high-dimensional space and how we do it has a direct impact on the system performance. We will discuss different techniques and what is new in that field which can make PostgreSQL the leading engine among vector databases.

In this presentation, we will discuss the various algorithms and indexing techniques for ANN vector search, examining tradeoffs between lookup speed, recall, memory consumption, index creation time, and other factors. We will also discuss the current state of vector search in PostgreSQL and some new ways we can contribute and help with development in that field.

Date:
2024 April 19 14:00 PDT
Duration:
50 min
Room:
San Pedro
Conference:
Postgres Conference 2024
Language:
English
Track:
Dev
Difficulty:
Intermediate