To automate the transformation of machine-readable PDFs to digital invoices, Basware has rolled out SmartPDF AI in the newest expansion of its SmartPDF offering. The tool uses information taken from past invoices through a cutting-edge machine learning (ML) model, according to a Tuesday (Sept. 22) announcement.
Basware said SmartPDF AI doesn’t need any installation, manual confirmation or configuration, while it offers support in different languages.
The self-learning system makes modifications on its own if a vendor sends a bill for the first time or if a current vendor adjusts its bill formats. The artificial intelligence (AI) technology considers features like logos and fonts instead of only the text, which lets it more effectively find patterns and bolster accuracy, according to the announcement.
Basware Director of Network Solutions Sami Nikula said in the announcement that the company had a vast array of information to educate the new AI models, “allowing us to confidently achieve the targeted field level accuracy of 97 percent.”
“Once applied to a customer’s historical invoices, the AI learns and improves even more each time, but again, with very little human interaction required, saving the accounts payable team a tremendous amount of time,” Nikula said.
Basware said it’s the only digital invoicing and procure-to-pay provider that offers companies complete spend visibility through full data capture and supplier connectivity. The firm trades on the Helsinki exchange and has offices in 14 nations.
The news comes as excelerateds2p, a consulting and implementation company, joined Basware’s international partner network, per news earlier this month. excelerateds2p has conventionally had a stronger focus on those who harness SAP Ariba and SAP.
In August, news surfaced that Sievo, a procurement analytics firm, had become a part of Basware’s partner ecosystem. Sievo, which was started in 2004 in Finland, lets firms look at their spending, supervise their contracts and make forecasts for their expenses, among other features.
Credit: Google News