Episode

Armchair Architects: LLMs & Vector Databases (Part 1)

with David Blank-Edelman, Uli Homann, Eric Charran

Azure Essentials Show

Vector databases are designed to store, manage, and index massive quantities of high-dimensional vector data efficiently that can help different types of queries, such as nearest neighbor. In this episode of the #AzureEnblementShow, Uli, Eric and David discuss how vector databases convert data to integers, cover some of the use cases of vector databases, and the benefits of embedding. This is part one of a two-part series.

Chapters

  • 00:00 - Introduction
  • 00:38 - Data stored as integers
  • 01:30 - Text converted to numerical data
  • 02:29 - Vectors are not new
  • 03:47 - Use cases
  • 05:02 - Benefits of Embedding
  • 07:40 - Vectorizing semantic concepts
  • 08:38 - Teaser for Part 2

Azure