@kmarrs901 the concept would work something like this (only a screenshot in the thread):
A sample workflow with a try-catch block that could be modified:
You could record which file failed. Ans also later use that collection to maybe process the failed files again.