The paperless office – a long-awaited utopia is increasingly becoming reality. We believe that the standardisation of all incoming channels, from physical incoming mail, to e-mails and fax, and through to digital communication via chat, social media and so on is a must!
An essential component of this is the automated assignment of document types or business transactions to interested parties, customers, dealers and other groups of people. But the more interesting part takes place behind the scenes. The background processing of subsequent processes can only succeed if all case-related information is extracted from the incoming business transactions. The information obtained for each business transaction is evaluated by our business process engine, enriched with further data, and then processed in the relevant systems on a case-by-case basis. Individual business transactions are only forwarded to experts via push routing or our frontend in exceptional circumstances. Background processing at its best!
Achieving top marks in automation and background processing depends on three essential success factors: understanding the process, knowledge of the incoming document types and business transactions, and the best possible technology for automating the OCR/ICR processing of the incoming information. At the latest, the understanding of the process comes about when the processes are graphically modelled via BPMN and when central questions are answered:
Which departments are affected? Which document types trigger which business transactions? What additional information is required from adjacent systems? And which information from the incoming objects is necessary? These are questions that always pursue the goal of merging all the information that enables a fully automated background case closure. Technology is the key to automation. Machine learning supports input management as the technological basis in the automated assignment of the document type and the identification of the business transaction. Complex search and extraction algorithms identify the information crucial to success for each incoming object and make it available to subsequent processes.