The Importance of a Vector Database

If you are working with data that isn't that doesn't need any semantic similarity, you don't need a vector database. Most models that we use to generate vector embeddings are trained on written text in the form of like sentences or paragraphs. And so if you have something like flavor colon, lime, comma, water colon still, it won't understand that as well as if you said something like this is a can of lime flavored sparkling water. It's not the right tool for that task.

Play episode from 33:12

Transcript

Episode notes

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app