The IDEAL Insurance Group is revo­lu­tio­ni­zing its input manage­ment through cutting-edge AI tech­no­lo­gies from Insiders. With the intro­duc­tion of the Cognitive Clas­si­fier and a deep learning-based OCR, the insurance company achieves a new level of quality. The con­clu­sive pro­ces­sing of straight­for­ward issues is now possible directly in the inbox – a quantum leap in effi­ci­ency.

Based in Berlin, the IDEAL Insurance Group offers its customers various products for securing biometric risks and has estab­lished itself as an innovator in the insurance industry with new concepts such as private care annuity or the first com­ple­tely digital insurance account „IDEAL Uni­vers­al­Life.“ The company always prio­ri­tizes the interests of its customers.

Since 2004, IDEAL has suc­cessfully utilized smart INVOICE for pro­ces­sing incoming invoices and later intro­duced smart FIX for automated incoming mail pro­ces­sing. IDEAL processes more than 500,000 documents annually, totaling around one million pages. The document intake is highly diverse, including hand­written letters, pho­to­gra­phed documents, faxes, emails with various attach­ments in different file formats, or embedded in the message, medical records, and official documents.

To minimize manual efforts, extract all relevant data accu­ra­tely, and promote automated pro­ces­sing with a focus on a high dark pro­ces­sing rate, IDEAL was in search of an advanced, intel­li­gent solution. This solution needed to handle poorly readable hand­written receipts, low image quality in pho­to­gra­phed documents, or the com­bi­na­tion of text and images on documents without any issues. A par­ti­cular challenge was the desire to con­clu­si­vely handle simple tasks – such as address changes or changes in bank details – already at the first contact with the document.

A clear case for the new Cognitive Clas­si­fier from Insiders! The Cognitive Clas­si­fier combines image-based clas­si­fi­ca­tion with optimized text-based clas­si­fi­ca­tion through advanced tech­no­lo­gies such as Natural Language Pro­ces­sing (NLP) and Deep Learning. With the reco­gni­tion of visual features, the Cognitive Clas­si­fier can now automate the reco­gni­tion and pro­ces­sing of items such as cer­ti­fi­cate rolls with seals, notarial documents, or photo attach­ments. Addi­tio­nally, the Universal Inte­gra­tion Layer for OCR (ILFO) allows the inte­gra­tion of various OCR com­pon­ents to address specific use cases. IDEAL also intro­duced a deep learning-based OCR along with the ILFO.

„Thanks to the Cognitive Clas­si­fier and ILFO, we have been able to crack the tough nuts in input manage­ment and make a signi­fi­cant leap in process auto­ma­tion. It’s great to be able to implement new AI tech­no­lo­gies in close col­la­bo­ra­tion with the manu­fac­turer,“ says Carsten Dickau, Business Analyst at the IDEAL Insurance Group.

Initially, the com­bi­na­tion of the Cognitive Clas­si­fier and deep learning-based OCR was tested on the par­ti­cu­larly chal­len­ging document class ‚Death Cer­ti­fi­cate.‘ This document class fre­quently appears in life insurance and serves as an example for all cer­ti­fi­cates whose extra­c­tion is par­ti­cu­larly chal­len­ging with a classic OCR due to printed water­marks. In the test, a signi­fi­cant quality leap was achieved: 97% of death cer­ti­fi­cates could be correctly automated, and in only 3 percent of cases, a verifier had to intervene.

This result thrilled IDEAL so much that the Cognitive Clas­si­fier and optimized OCR were intro­duced for all document classes of the insurer. By combining image and text infor­ma­tion, the Cognitive Clas­si­fier excels even with pre­viously hard-to-classify document classes, incre­asing reco­gni­tion rates by another appro­xi­m­ately 15 percent to nearly 100 percent. The alter­na­tive OCR connected via ILFO also con­tri­butes to improved reco­gni­tion in the case of qua­li­ta­tively poor photos, faxes, or scans, as well as hand­written documents, signi­fi­cantly enhancing the higher level of auto­ma­tion.

With these inno­va­tive AI com­pon­ents, IDEAL can now process infor­ma­tion that was pre­viously difficult or impos­sible to extract. The con­clu­sive pro­ces­sing of simple tasks in input manage­ment has also been imple­mented.

The IDEAL example demons­trates that providing a platform where various AI tech­no­lo­gies can col­la­bo­rate con­sis­t­ently brings process impro­ve­ments for customers through tech­no­lo­gical advance­ments.