Hi @wskinnermctc and @Nived_Nambiar
So, the flow does several things. When a file is dropped into the library, the AI model engages and updates the file metadata based on the file content. There's a feedback loop as well where if the confidence score of a given category is too low, the file gets stored for re-tagging. Unfortunately, the first way I designed the feedback loop, it worked but instead of the actual file getting stored for re-tagging, just the data of the file got stored-- and since it wasn't a PDF, I couldn't actually tag it or do anything with it within the AI model. The apply to each was me trying to create an attachment variable for the flow to store.
I have been trying to solve this problem (the problem of putting the actual PDF into storage for re-tagging) for weeks with no success, and now the flow itself isn't working because of the changes I've made to try getting this PDF attachment stored correctly. I hope that makes sense. Anyway here is the flow in its current state:
It's a very long flow and I have klodged it together-- I seriously doubt that all of the steps are necessary, but I don't know which ones to remove and I didn't want to break the flow altogether.
