Is the Typesense wordpress plugin able to support the newly released vector search feature? Eg. Get data from wordpress backend -> pass the data to an embedding model to generate the embeddings -> import the embeddings back into typesense -> to do vector search?

cc @Digamber Pradhan

Not At the moment, no.

@Digamber Pradhan This is for a project I'm working on. We are looking to host a use case search engine on Wordpress using Typesense. End users (SMEs) can have access to the use case information by inputting search query to find the matching use cases and the relevant solution providers. There'll be separate phases to the project. For first phase, we'll be using Typesense WordPress plugin to implement the default search. Subsequently for the next phase, we're planning to introduce semantic search. I'm trying to explore what's the best way to ingest the embeddings into Typesense. Since for the first phase, we have decided to go ahead with using the Typesense Wordpress plugin. But for the next phase, I'm looking to see if the plugin is able to support else might have to figure a workaround.

Ah thanks for the explanation, there isn’t an out of the box solution for this right now. From what you’re telling me here - you probably could use the plugin to act as the interface and change / extend the underlying functionality for Vector Search - but as I haven’t implemented it myself I can’t say for sure.

@Digamber Pradhan Anyway, we've just tried testing the plugin for custom post types. Submitted a dummy form but it seems like the contents are indexed but the image is not because the thumbnail url is not updated. Is there a way for the image to be reflected on the thumbnail?

Would be determine by your code. Are you adding the post thumbnail as a meta field ? Or as the post thumbnail? Can you provide me code for your implementation ?