Just for reference, on page 31 at the bottom is some text "Reference ID: 3610837" It is not in the json output from the extraction API. I have attached the original PDF as well as the json output of your tool.
Not that it's a showstopper for us, but if that is the confirmed reason that the text is not being extracted from the footer, is there a chance in a future sprint/improvement cycle of the tool that a config option can be added to broaden the text search to the entire page?