While the term “automation” has taken the Accounts Payable world by storm in the last few years, now the phrases “machine learning” and “Artificial Intelligence” are beginning to crop up more often. While these phrases may bring to mind movies in which rebellious robot armies turn on their hapless human creators, the reality is quite different. In the next few years, most finance departments will find themselves dealing with some sort of machine learning component in their software solutions. Even Forbes weighed in on the topic:
“It is high time for every accountant to reflect on their job, identify the opportunities machine learning could offer to them, and focus less on the tarks[sic] that can be automated and more on those inherently human aspects of their jobs.”
As part of a series into how machine learning will affect AP, we’ll look at how this new technology differs from automation.
AP Automation: Brawn without Brains
Automation has already had a strong impact on the role of Accounts Payable. Regimented approval workflows, Optical Character Recognition (OCR) technology, and automated invoice processing have removed many manual aspects of the job. However, though automation can improve AP efficiency, they are ultimately static solutions. Once implementation is complete, the workflows are in place, and AP staff are trained on the process, the solution has reached its highest potential for improving business operations.
How is Machine Learning Like Your Favorite Barista?
Just like the barista at your favorite cafe, Accounts Payable software that incorporates machine learning gets to know you, gets to know your business, and serves up your “brew” just the way you need it!
An Accounts Payable solution that incorporates machine learning will constantly look for ways to make your job easier. Data gathered from daily operations can be analyzed and leveraged to make the Accounts Payable process faster. For example, GL code entry. Machine learning capabilities allow the system to recognize patterns, such as when a certain GL code is applied to a certain vendor. Over time, the machine ‘learns’ these patterns, and can suggest this GL code whenever an invoice from that vendor is being processed. And that’s just the tip of the iceberg. This “Artificial Intelligence” process happens with almost no extra effort on the part of the Accounts Payable employee. The system’s initial learning curve is quickly balanced out by smoother, faster invoice processing.
To return to the Forbes reference above, this allows Accounts Payable to turn their attention to the more human aspects of their job – Invoice Approvals. This is where Accounts Payable gets interesting. In our next blog we’ll explore how machine learning coupled with a dynamic approval workflow can help Accounts Payable reach its potential as a critical part of your finance team.
Curious about what an Artificial Intelligence component looks like? Meet Billy the Bot, Stampli’s machine learning assistant. He likes to hang out in invoice fields where he can help find the right codes faster. Schedule a brief demo to watch him in action!