The customer is a global insurer and financial services provider with over 8,000 employees and revenue of over $530 million. Its mission is to make decisions easier and the lives of its customers better.
The insurer’s Long-Term Care division handled bank authorisation charge processing manually. The large volume of documents received contained vast amount of unstructured data, including handwriting. Each of these documents required reading and manual entry into the company’s downstream systems while signatures also needed checking.
CMR improved the document quality with its pre-processing engine and document indexing. Handwritten fields and checkboxes were dynamically extracted and the presence of a signature verified using advanced Deep Learning techniques. Exceptions were handled using the innovative quality check screen, enabling CMR to learn with both assistive and adaptive Machine Learning. Clean data was fed directly into the customer’s downstream systems using API integrations
85-92% accuracy for handwriting recognition, enabling processing of unstructured data
75% reduction in manual work for tedious process of data extraction
This insurer now experiences higher business productivity, faster turn-around time, increased accuracy and better workflow management.