Xano R1.59 Release Announcement

0:00 Intro
0:14 Realtime
0:47 Database Vector Support
1:02 New Onboarding Experience
1:21 Internal Documentation Tools
Full features, bugs, and fixes list: community.xano.com/release-an...
Get a deeper dive into Realtime: • Realtime (Websockets) ...
Explore Vector Embeddings: • Building AI and ML app...
Xano - The Fastest No Code Backend Dev Platform
go.xano.co/3K0uE8P
Twitter
/ nocodebackend
Subscribe to Xano's Channel for Weekly New Content
/ nocodebackend

[@TechAuditTV]

Xano woke up and decided to drop the iPhone 16, 17 and 18 all in one release 🤯

[@pedrogermano15]

Game changer! Congrats, Xano team!!!

[@brunnononato3534]

Good

[@alexander_world]

I have a table of 4000 embeddings (1536 dimensions), correctly indexed on Inner Product and plain query takes almost 10 seconds. Xano support just tells me: "I have not used pgvector so I'm not sure about that. I do know that storing such large arrays in the database would slow it down exponentially, so you would have to store it elsewhere and then query for them."
Sounds like it's just unusable then...

[@LachlanXano]

Hey there, there is actually a unique aspect of the vector indexes with PGvector that requires your Inner Product queries to also contain a ascending sort in order for the index to apply. This is why for similarity search, where you are looking to find the most similar items we recommend using a Negative Inner Product query with a ascending sort. If you would like to return the least similar you can use a inner product query with an ascending sort. The inner product index works for both Negative and standard inner product queries. Please give this a try and you'll find that your queries will be very fast.