Is there a way to upload data to n8n for an LLM to reference with a prompt or do I need to include all the data in the prompt itself?
Attach a vector store & embedding model. This way the AI agent will search through your documents/data and retrieve chunks as context.
Is this something native to the platform or do I need to sign up with a third party to accomplish this?
You do need credentials, like any other node in n8n. Look up the documentation about vector stores.
Appreciate you
1 Like