Artificial intelligence (AI) is everywhere. Rudy Lauwereins, professor at the University in Leuven made that clear once again at EIT Digital Innovation Day last December. AI systems are the future. But what does that actually mean for supporting procurement and processing incoming invoices?
By Vincent Wouters
Everyone has worked with formulas, which is already AI in the basics. That was this Belgian professor's starting point. But as soon as the formulas become more complex and the amount of data increases, it stops for a normal human being. However, computers can handle this just fine; Moore's Law helped make this happen.
This means that by now we are able to develop very smart applications with AI. Think automated medical diagnoses based on image recognition, personalized Spotify playlists based on large amounts of user data to self-driving cars via sensors and cameras.
Virtual assistant
Suppliers of purchase-to-pay solutions(think procurement support and invoice processing systems) are also using AI in different ways. After all, there is still much to be gained in procurement and finance departments. For example, several parties launched their own chatbot feature. A virtual assistant that helps users place orders and search for items or orders.
The chatbots combine natural language processing technology with AI to offer users a simplified way to interact with the system. They can interact with the assistant as they would with a human.
In addition, the system learns by combining search results with the final order. For example, when searched for "personal computer," the system learns to show laptop systems (as well). This greatly improves the ease of use of the system. It also reduces training time for new users.
Booking proposals
But AI goes beyond the virtual assistant. Case in point, the finance department is an extremely data-rich part of the organization. Based on these large amounts of data, the software will soon be able to predict things with great accuracy.
The better solutions currently already offer functions to also automatically book, for example, the non-order related (non-PO) invoices, the so-called smart coding technology. This uses large amounts of historical invoices and automatically applies the most similar coding structure to the new invoice. The automatic selection of the correct procuration scheme can also be done quite reliably.
In addition, systems are capable of creating purchase orders themselves based on historical data. Or to bundle requisitions. In this way, organizations can save on transportation costs, or take advantage of volume discounts.
Longing perspective
In the combination of AI with reporting, I see opportunities for further optimizing allowable deviations in automatic invoice matching. Often simple thresholding rules can then be replaced by rules that further optimize the financial process and reduce costs. The basis for these smart and learning systems is so-called "corporate spend data.
Furthermore, it is important to use platforms that are constantly evolving with the technologies mentioned earlier. All this will bring next-generation solutions for further optimization of the source-to-pay process. The benefits are legion: better compliance, improved business performance, reduced procurement costs, suppliers delivering at the right time, et cetera.
We have since moved to digital invoices, order-based purchasing with automatic invoice "matching" and the use of predefined payment plans, for example. Meanwhile, the technology of AI and machine learning is far enough along to take the next step. A beautiful beckoning perspective for procurement and finance!
Never miss a blog again?
Sign up for the newsletter and receive the latest blogs automatically in your mailbox every month.
