When it comes to invoices, it’s all in the eye of the beholder to determine which amounts are due for what services or goods. If an accounts payable clerk receives a document, whether by snail mail, fax or as an email attachment, they have to use their common sense and experience to identify and manually enter the relevant information into the accounting system. It takes time, but it usually works and is occasionally plagued by entry errors or misunderstandings, say from a quirky invoice format or bad handwriting. That way of literally “eyeballing” it, though, is quickly becoming a thing of the past.
Systems for AP Automation have made ingesting invoices so much easier, it’s worth taking a closer look at how we at Yooz have built a cloud-based platform that is faster and more accurate than an experienced human.
It starts with optical character recognition (OCR). OCR is the mechanical or electronic conversion of images of typed, printed, or even handwritten text, into machine-encoded text. The text can come from a scanned document, a photo of a document, a scene-photo (for example writing on signs and billboards in a photo), or from subtitles superimposed on an image (such as a TV show). It’s a nifty shortcut to save time typing, but it’s only the first step.
The next layer -- smart data extraction -- is where the magic starts to happen. Smart data extraction technology basically “reads” each invoice. It can understand and process the text from the OCR and transform it into relevant data. OCR by itself does not know what to do with the information it sees, but the “smart” in smart data extraction adds a crucial capability: software can now take the text and do something with it, instantly making the life of every AP clerk so much easier.
With Yooz, the relevant fields in an invoice such as the date, amounts and applicable taxes, discounts, due date and recipient information plus payment info are automatically moved into the corresponding fields of your accounting or ERP system’s database. The software assigns the correct GL code, matches purchasing orders to invoices and will move the document along to the right person to approve it. Or flag it if it’s a duplicate or otherwise problematic. At its best, this is a touchfree workflow from end to end that cuts the time to process an invoice down to minutes and a few cents.
But state-of-art AP Automation doesn’t stop there, thanks to the next layer called machine learning, which is a subset of artificial intelligence. Simply put, it means that Yooz remembers the data it has seen and will keep populating it into the correct data fields each time it sees this type of information again. In short, it’s hard to stump the system.
No matter how different invoices from supplier A to supplier Z may look or are formatted, machine learning gets each variation and reduces errors. More importantly, it learns over time and improves. No end-user is ever involved to “teach” the software. Instead, your back-office staff is freed from both manual data entry and third-party verification. And since Yooz has been working on this intelligent AP Automation stack for over a decade and “read” millions of invoices, we continuously teach our software new tricks that instantly benefit every one of our users.
While all this decoding of business transactions is happening thousands of times every day, it does more than just save time and money. AP Automation gives companies of any size the big picture. Understanding what flows through your AP department almost in real time provides the knowledge to better manage your working capital, mitigate risk and make more strategic decisions.
We think that’s exactly the kind of thing humans deserve recognition for.