The only constraint of the service is that the PDF must be machine generated, i.e. produced directly from the billing application raising the invoice.
When a PDF is machine generated in almost all cases it will be a text PDF, with the invoice data items carried directly within the PDF. Our service takes advantage of this and maps the invoice data direct from the PDF to an e-invoice structure – in other words, the service is akin to mapping one flavour of XML to another flavour of XML.
If the PDF is generated by scanning the invoice, an image or photo of the invoice will be passed in the PDF. Unfortunately the only way to process image PDFs is to use OCR (Optical Character Recognition) technology – this approach cannot guarantee data quality. For these reasons we are not able to accept image PDF files.