Did you know you could use vector database indexes with Postgres?

All you need is the PG vector extension for Postgres.

If you're not familiar with vector DBs and you want me to do a deeper dive on them, feel free to shoot me a message, comment, etc. I'd love to do a deeper dive on this if you're interested.

Now you know me, I'm all about building things at scale and helping people scale up. So my first question was, does it support sharding/partitioning? The good news is it does!

CREATE TABLE items (embedding vector(3), category_id int) PARTITION BY LIST(category_id);

Now, I don't think you could use the vector as a partition key. It looks like they prefer you to use something more standard, like an integer. But it's cool to know you could partition it all the same and that index could work across multiple partitions.

If you're interested in learning more about how to scale your vector DB to handle massive workloads from your LLM agents, then you should check out my Tech Talk this Friday at 1PM CT.