How does train from URL work?

Does it work by actually crawling those pages or does it ask google for search results restricted to those pages? Or something else entirely? I doubt when I told it I wanted to search SEC.gov that it went and downloaded all those pages… How can I see whats happening under the hood?

Hi @ryan1 ,

We use Bing to search the website. Those two live sessions might hold useful information for you:

@bassam.tantawi

We’d like to store vector data with a db like Pinecone https://www.pinecone.io/

Is that possible with BP Cloud? thx

Hi @pingriser ,

You can use an Execute Code card to call it and retrieve data