Basware Deploys Machine Learning For Expanded Invoice Automation

Source-to-pay (S2P) company Basware is tackling the issue of manual coding of invoices not linked to purchase orders.

The company announced news on Thursday (Nov. 9) that it is deploying a machine learning-based tool to automatically code non-PO invoices. The “smart coding” capability, as Basware described it, identifies where costs should be allocated and accurately codes the invoice.

“Traditionally, non-PO invoices are automated with business rules that are laborious for customers to manage,” explained Basware VP of Purchase-to-Pay Product Management, Sami Peltonen, in a statement. “To create a more touchless process, we developed an intelligent algorithm to determine the most accurate cost allocation for non-PO invoices.”

“The solution uses machine learning to analyze customers’ historical transactions and then provides a highly reliable recommendation to the user,” Peltonen continued.

Without automated cost allocation identification, Basware said employees can struggle to correctly process invoices, especially those not familiar with bookkeeping rules.

Basware noted that the only manual intervention necessary with its new tool is for professionals to check that the coding is correct.

“Through the years, we’ve worked closely with our customers to simplify their operations and help them spend smarter through forward-thinking approaches to automating finance and procurement processes,” added Basware SVP of Purchase-to-Pay Business Area, Ilari Nurmi, in another statement. “Our customers are always inspiring us to improve and deliver innovations that drive measurable value for their organizations, and smart coding is a direct result of that.”

The launch of its smart coding solution comes a year after the S2P company announced the creation of its procurement Marketplace, an online portal through which businesses can procure goods and manage supplier relationships.