Once you retrieve the parsed documents from llamaparse, create an index and store it somewhere.
So for next time you can check first whether you already have the index persisted with you or not. If not then proceed with parsing else load it from the stored place.
