Wow, what good news! I totally love Hasura but totally missed a lot of things since graphql. My stand was, that i don't want to use a second product for vectors and i'm totally into postgres and hasura.
My plan was to add: pgai
I came along a blog post from timescale last week, it's worth a read: https://www.timescale.com/blog/vector-databases-are-the-wron...
My question: do the newly added hasura features make it completely obsolete to add pgai as extension to postgres or can they play along nicely.
My first guess is i'll simply combine both. I want to have a stack i can work and get deeply fimilar with and not change it all 3 month, when a new funky vectordb is announced.
Drop the "GraphQL" -- it's cooler.
I have built and managed more 10 production codebases that use Hasura -- IMO, I think it's a better option than Supabase, Payload, and others to build your backend. The fact that they are investing time in LLM-related resources for the platform makes me happy we choosed it back when.
LLM's and a framework that's structured and clear enough for LLM's to understand (if you have used Payload CMS and built blocks, fields, or other, this is also an example) has droven a lot of growth in my company. People that took more than 6 months to be productive in our codebases are becoming much more productive way sooner -- and this is not only because Sonnet and 4o are so good, but because we also have tools that put effort in to making sure LLM understand them, and Hasura is one of them.