The documentation seems like it's focused on producing aggregate data but is there a simple way I can just give it a single document of text and have it spit out what it thinks the location is?