I’m building out and customizing the workflow from Cole’s “Ultimate n8n Agentic RAG Template” and having HUGE issues vectorizing large, complex pdf files with the default data loader and embedding in that workflow. So I’m looking at using Vectorize.io or possibly Unstract to build a ROBUST RAG workflow in n8n that can properly vectorize complex pdf files.
Any thoughts, help or tutorial links would be appreciated!
I haven’t used Vectorize myself but I might be making a video in the future on creating a RAG pipeline with Unstract! Nothing concrete I could share at this point though.
Interesting. I just had a webinar with the engineers at Unstract and it’s a very powerful engine. We provided an image of a damaged roof with very hard to read text and hand written scribbles overlaid in different positions and Unstract read it in and produced a clean formatted, easy to read table.
It seems like Unstract could be a solution for our pipeline, I’m excited to see what you come up with!
That’s awesome, love it!
