Sramana Mitra: What is an example of what you’re trying to do with images? What specifically are you trying to achieve with these images?
Shan Haq: One of the things that is important to our customers is the ability to bring a supplier onto our network very quickly. The most common scenario there is being able to take an invoice format from a supplier and turn it into the appropriate invoice format for our customer.
One of the things that we need to do is build that business rule around the invoice. If we can get the invoice from a supplier and very quickly identify, just from the image, that that invoice is coming from a Quickbooks system for example, we can put it into a subgroup and we know the type of business rule that we’re going to need to construct for that particular system.
If we can look at the image of the invoice and see that this one came from Oracle, then we can put it into a different subgroup. It can expedite the process and make it much more efficient in terms of getting that supplier invoice into the system.
Sramana Mitra: Is there any other use case that is worth discussing in the context of AI conversation?
Shan Haq: One area would be predictive analytics. What our customers want is not only an electronic invoice but also the ability to do things with a straight-through-process. They get an invoice into their AP system from a supplier. They want to be able to process it without a human touchpoint. One of the challenges is that when a supplier sends an invoice and they do things like put the PO line out of number, that makes it difficult for a customer to match that invoice to the purchase order.
We leverage AI through a predictive analytics standpoint because we have technology that when the supplier doesn’t put the PO number on the invoice, we have the ability to predict that information based on the PO and invoice data. This results in a much better scenario for our customer because when a supplier sends an invoice, we can populate even though the supplier didn’t provide the information.
Sramana Mitra: Great. Go out a little bit and look at the AI industry in general. Talk to me about what your observations are about open problems in the space. Where do you see new AI company-building opportunities?
Shan Haq: I think there’re two broad areas that I see out there. One is, there is a lot of data. It’s not so much that we can’t get the data in order to make better decisions and communicate more collaboratively. It’s how we manage it. How can you consume the data in a way that you can organize it and make it practical?
We’re a part of trying to solve some of those problems. We’ve got suppliers connected to our network. They are sending transactions through our network. We know something about the suppliers because we provision them before we put them on the network. Just being able to capture all of that and providing that information to customers is a problem that companies can help solve.