While all invoices include similar fields, they are sent to AP operations in hundreds, sometimes thousands, of different formats from different suppliers. The work of extracting data from each invoice and entering it into applications like SAP is repetitive, but variability of the formats makes it very difficult and expensive to program “if-then- else” rules in RPA. The frequency of changes to these documents makes a template-based automation approach even more difficult, and cloud-based AI solutions are too expensive and require data science expertise. Additionally, manual processing is time- and labor-intensive, and exceptions must be handled on a case-by-case basis.
WorkFusion SPA provides reusable components that AP teams can quickly configure, without IT support, to automate their unique invoice- handling process. OCR captures and digitizes text, and SPA creates cognitive bots to extract key data with its patented Process AutoML, which uses historical and real-time human work to train machine learning models. SPA automatically flags missing or unreadable information as exceptions and routes tasks to the right people. Each manual intervention improves the completeness and confidence of machine learning models, which makes cognitive bots smarter and more accurate. Finally, RPA bots send validated payment requests to Accounts Payable for processing. These same capabilities can be used to extract and structure data from any type of document, such as bills of lading, hospital bills, credit letters and mortgage deeds.