Do I really need to create vector search for every collection?

Couchbase Capella
, , ,
1

Hi everyone!

I am very new to Couchbase and databases in general. I am using the Couchbase SDK.

I am trying to make an RAG-based LLM to track social media analytics.

I wish to input a string, which can be a person describing his social media post idea, and the RAG model returns the likelihood of going viral.

Here is my database setup:

social-analytics (BUCKET)
│
├── content (SCOPE)
│ ├── posts
│ ├── topics
│ ├── behaviors
│ └── events
│
├── lexicon (SCOPE)
│ ├── emoji
│ ├── hashtags
│ ├── slang
│ └── keywords
│
└── activity (SCOPE)
└── logs

I noticed that I have to create every single vector index for each scope and collection manually Surely there must be a more compact way to do this?

When the user enters a query I need to be able to search between scopes because a user can post a combination of emojis and topics

Thanks,

Martin

3

Sorry for the late reply! You can create Search service vector indexes for multiple collections per scope, so that would lower the number of vector indexes you’ll have to create. You can’t create one covering more than one scope, however.

1 Like