PlanetScale, the cloud MySQL-compatible database with advanced scaling capabilities, is now upgraded with vector storage and search.

PlanetScale's main selling points include horizontal sharding and non-blocking schema changes. Adding vector capabilities into the mix, makes it an even more robust and flexible database solution. 

As a refresher, vector embeddings take a piece of text and create a numerical representation of that text. Text with similar content will have similar vectors in this numeric space. This allows us to compare those vectors and find pieces of text that are similar.

Saying that, there's two aspects in adding vector support to a database; storing vectors and then using them for searching. Usually storing vectors requires a dedicated database like ChromaDB. Not anymore, as we examined in "Turn PostgreSQL Into A Vector Store", the pg_vector extension that renders Postgres as a viable alternative to specialized vector stores totally replacing them. The big deal now is that you can seamlessly store your vector data alongside your application's relational data.

This capability then enables you to do operations like semantic search, recommendations or anomaly detection, faster and more easily than ever before - and without extensive application changes.

PlanetScale is now doing the same but for MySQL, with a twist.
While the implementation supports the following out of the box:

• Pre-filtering and post-filtering
  
  
• Full SQL syntax — including JOIN, WHERE, and subqueries
  
  
• ACID compliance

it does so by using a new and more advanced vector search algorithm by extending the SPFresh one of Microsoft,
which allegedly performs better than HNSW and DiskANN.
That is so because the aforementioned algorithms require periodically re-building of the index with the underlying vector data, SPFresh provides real-Time Synchronization of vector embeddings with the latest data changes; therefore avoiding staleness with the added bonus of transactional support on top of it. This property allows for developers to use Mysql as their mani tool for building AI powered applications.

To enable the vector support (in beta at the time of writing) in your own managed instance, you have to, from the main dashboard, navigate to the database settings page, then click "Beta features", find "Vectors" and click "Enroll". And with that you get a scalable, reliable and permanent database with full vector capabilities.

More Information

Announcing the PlanetScale vectors public beta 

Related Articles

Turn PostgreSQL Into A Vector Store  

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook or Linkedin.

Comments

Make a Comment or View Existing Comments Using Disqus

or email your comment to: comments@i-programmer.info