What’s happening in the Visma Machine Learning team?

If you believe the Gartner Group, AI and machine learning are at the peak of the hype cycle these days. Inside Visma, the story is a little bit different. We’re busy making machine learning an everyday productivity boost for our customers – and we’ve been doing it for years.

As technology matures, some of the initial excitement might die off, but only because the technology starts to appear everywhere. We are moving from the hype phase to the install phase. Instead of talking about new technologies we start using them.

For the Visma e-conomic Machine Learning team – the VML team for short – this means providing task automation for thousands of companies every month. We’ve had an exciting couple of years inventing these new technologies – and now we’re making use of them. We’ve harnessed our machine learning know-how into two solid machine learning engines, and are rapidly adding the advantage of machine learning to more and more of Visma’s products.

How does this work for our customers?

When you upload the image of your travel expense receipts using Visma Attach, and don’t have to type in the amount and date from the receipt, you’re using the Smartscan engine. Developed by the VML team, SmartScan reads and understands the information on the receipt.

When you process your supplier invoices through e-conomic’s Smart Inbox and the invoice is automatically booked, you are combining Smartscan with the AutoSuggest engine, which learns exactly how you process your invoices and do it for you.

During 2018 you will be finding these simplified, automated workflows in more and more Visma products. In addition to that, we’re constantly working with all the other software development teams in Visma to find new uses for our machine learning engines. Task automation is just the first step for Smartscan and AutoSuggest.