I would like to finalize the process of uploading data to a vector database as follows: the user sends a PDF file to the chat, which should be the trigger for activating the process; then this file should be uploaded to Pinecone. At the moment, I am hosting the file on google disk, and running the process for each file separately.
Hey @Egor12341234 hope all is well.
If you need the user to provide the PDF to a chat, you need to decide which chat you want to use - embedded n8n chat (local or integrated into a website) or Telegram or Slack. Do you have anything specific in mind for the trigger product?
let’s look at the option of sending via telegram bot, it is important for me to understand which nodes are between the TG bot trigger → ??? → pinecone
In most basic setup, you would generally (given you PDF is a text pdf) have extract data node to get the content, then you send it to the ai agent for data loading / splitting.
