FileByMagic is a revolutionary data capture and document filing system from LandoByte that automates these tasks, saving companies time and money, and allowing them to focus on their business.
Machine based data mining from documentation is not a new idea. It has been around in various forms for decades but has always been difficult to achieve outside of very controlled environments because of the variability in documentation. Even among relatively standardised documents such as invoices the difference between any given two can be astounding.
So what is it that makes understanding documentation something that humans do naturally and machines fail at so regularly? The answer is context.
When a human picks up a piece of paper the first thing we do is establish what type of document it is. Invoice? Bank statement? Contract? Party invitation? There are clues that we look for in each of these and once we’ve picked them up we know how to handle the document. You don’t look for a RSVP date on a bank statement or account balance on a party invitation.
FileByMagic was built from the ground up based on these principles and in its development the question we continuously return to has been: How would a human do this?
By first establishing the type of document, the system is able to determine what information it will be able to obtain. It then uses clues on the document such as the presence of logos and the contents of footers to determine who produced the document, which in turn gives it more information on how to find the information it needs.
At the heart of FileByMagic is a unique artificial intelligence engine that is constantly learning and expanding its intelligence. It is backed by a suite of over a hundred rules that is growing continuously.