OCR Scanning to our database

We basically need to write code which can poll our ftp location every few minutes and as soon there is a PDF file (our DHL invoice or other courier's invoice) then it will pick up this file and use the OCR function to get the text out and then parse this text correctly and enter this information in our database. So basically we will have complete record of this invoice in our own format like tracking ID, shipment name, customer name, destination country, destination city, destination post code and the cost of that shipment and any fuel surcharge attached to that shipment. We have invoices in PDF coming from our couriers like UPS and DHL and they each have their own formats and structure and with the text data, we should be able to parse and store this information in database. If you can do good work, there are tons of other things attached to this same project later which I'll pass on to you. Thanks Sunny