"Basware algorithm-based machine learning services are available to help add value at every point in the purchase-to-pay process by freeing up employee time, eliminating errors, and speeding things up," said Sami Peltonen, VP Products and Business Management of North America for Basware. "Earlier this year, we introduced a natural language processing, artificial intelligence based virtual assistant to help users navigate our electronic procurement solution and find transactions much faster. Another capability we just introduced is what we call intelligent order aggregation."

"The system analyzes historical data to automatically determine the number of approved items waiting to be ordered and compares it to the weekly average for that particular vendor and location. If the number of pending items meets or exceeds the average, a purchase order can be automatically generated. If it is less than average, however, the user has the option to wait until the average number of items is reached before placing the order so that they can save on freight costs as well as take advantage of any volume discounts offered by the vendor," continued Peltonen. "We know that buyers would benefit from a more efficient way to combine requests without interrupting the flow of business. It was a perfect opportunity to apply machine learning."

The other areas where Basware has introduced machine learning based functionality include product mapping and catalog search. One of the biggest challenges for buying organizations is that the product codes they use internally can be very different from what suppliers use on their end. In this scenario, machine learning can fill the gaps. The new Basware product mapping feature organizes products into the right category, so that suppliers simply upload their catalogs and products which are then automatically matched to the buyer’s defined category, making the catalog management process easier for both buyers and suppliers.

On the catalog search side, the Basware electronic procurement solution leverages a flexible product master data model behind the scenes, so that procurement departments can create a database of products with multiple suppliers, prices, units, etc. for each product where needed. Then a highly intelligent algorithm automatically selects or suggests the best option based on pre-defined criteria like availability, location, price or delivery schedules, without exposing end users to all the complexity in the back-end.

For the end user, search query optimization makes the process of finding and ordering products and services very easy. The Basware solution creates a link between queries that come up empty handed and the items that are then added to the shopping cart. In other words, if a user searches for “personal computer” which yields no results, but then later adds a “laptop” to the shopping cart, the solution learns from these actions and gains a better understanding of what the user was looking for to begin with. Pairing together empty search results with the eventual shopping cart addition minimizes fruitless searches. This kind of search functionality promotes efficiency and makes ordering products simple for the end user.

"Embracing machine learning and artificial intelligence can add tremendous value to both buying organizations and suppliers," concluded Peltonen. "Our customers typically manage 100% of their spend through our platform. That puts Basware in a unique opportunity to leverage all the data for further innovation."