Hi, With the help of claude, I’ve already built a working RAG AI Help Agent for my business. My workflow includes retrieving relevant information from a pinecone vector store, then passing it to a code node which will then match the query with the highest score or relevance from the vector store and generate the chat prompt for the AI Agent node. The response it gives is almost 90% accurate, So I’m still trying to tweak it.
But I was wondering if this is the best approach or is it better to add the pinecone vector store as a tool sub-node on the AI Agent node? Then just generate the chat prompt from the option as a Human Message and System Message? Thank you.
Information on your n8n setup
- n8n version: 1.100.1
- Database: PostgreSQL
- n8n EXECUTIONS_PROCESS setting (default: own, main): Own
- Running n8n via (Docker, npm, n8n cloud, desktop app): Docker
- Operating system: Linux Ubuntu