When an agent uploads a big pdf that contains several documents, we should have the ability to split the big PDF so we can attached individual documents to individual tasks.