Some years ago, one of our customers, a large multinational agricultural company, decided to tackle the many document data capture challenges at the source by automating the extraction and classification of inbound documents, in particular purchase invoices. The company wanted to improve the document throughput and automate processes to reduce manual labour and improve productivity. They approached PacSol to recommend a solution to underpin the process and so we helped to implement IBM Datacap. Our biggest challenge was to roll out the capture solution to multiple countries (initially mainly in Europe) and therefore to process documents that included text data in many different languages. Previously the majority of the documents were manually sorted and scanned using paper copies but now by far the majority are received electronically via emails and the PacSol tools are used to feed the documents to Datacap. In addition to all this we have also helped to integrate Datacap with PacSol’s IBM i based ECM system & ERP system to allow them to fulfil their SLA agreements with their customers and suppliers on the payments terms.
IBM Datacap includes natural language processing and machine-learning technologies to help capture more unstructured as well as structured content and this facilitates the data capture from a wide variety of documents.