Vector indexing approaches explained

Nik and Michael are joined by Simon Eskildsen from turbopuffer — among other things, they discuss ANN index types, tradeoffs that can make sense for search workloads, and when it can make sense to move search out of Postgres.
 
Here are some links to things they mentioned:

• Simon Eskildsen https://postgres.fm/people/simon-eskildsen
• turbopuffer https://turbopuffer.com
• Use ULID Idempotency Keys (tip 6 in this blog post from Shopify) https://shopify.engineering/building-resilient-payment-systems
• PostgreSQL 18 Release Candidate 1 https://www.postgresql.org/about/news/postgresql-18-rc-1-released-3130
• Understanding DiskANN (blog post by Junaid Ahmed) https://www.tigerdata.com/blog/understanding-diskann
• SPFresh: Incremental In-Place Update for Billion-Scale Vector Search (paper) https://arxiv.org/abs/2410.14452
• Amazon S3 adds new functionality for conditional writes https://aws.amazon.com/about-aws/whats-new/2024/11/amazon-s3-functionality-conditional-writes
• Amazon S3 Vectors https://aws.amazon.com/s3/features/vectors

~~~

What did you like or not like? What should we discuss next time? Let us know via a YouTube comment, on social media, or by commenting on our Google doc!

~~~

Postgres FM is produced by:

• Michael Christofides, founder of pgMustard
• Nikolay Samokhvalov, founder of Postgres.ai

With credit to:

• Jessie Draws for the elephant artwork