I'm planning to support exactly that workflow.

Drop a bunch of PDFs on the program, queue them up for (later) processing, (auto)-tag, categorize and fulltext index them. I'd be so happy when that's done