1B row vector search in less than a second with Azure SQL Database Hyperscale | Data Exposed

Why watch

This video demonstrates Azure SQL Database Hyperscale's ability to perform vector search on 1 billion rows in milliseconds using DiskANN, a graph-based vector index. It covers the engineering decisions behind the index, shows demos of iterative filtering and DML operations on vectors, and explains how the query optimizer handles approximate vs exact vector search. Ideal for developers building AI/ML applications who need vector similarity search in a managed SQL database.