Users should be able to upload folders (consisting of 1 or more files and/or folders) or multiple-files directly to a cloud storage bucket via Data Explorer, and then use those folders (or files) as input (or output) to a pipeline run.