Live Demo

Claims Indexation

A Fortune 500 insurer wanted to reduce inefficiencies and errors caused by manual document classification and data extraction. omni:us automated the indexation and processing of 100.000 unstructured workers’ compensation claims notification documents provided per day via physical documents, letters, and e-mails.

Warning-2

Issue

  1. Ineffective processes and lack of technology applied demands a highly manual and personnel intensive claims file creation process
  2. Complex business rules on document classifications causing false routing and processing loops
  3. Insufficient and manual data structuring procedures frequently cause errors during data extraction resulting in additional processing loops
  4. Incomplete manual data capturing after document intake results in a significant data loss between indexing (registration of claim) and adjustment (claim handling) again increasing processing loops
Target

Impact

  1. Decreased average claims processing time
  2. Reduced complexity of business rules for document classes from 32 to 16 becoming MECE and eliminating redundancies
  3. Lowered amount of document misclassifications 
  4. Redeployed employees from manual steps to sophisticated work  
Clipboard-check

Solution

  1. Standardized document classification guidelines and redefined relevant data extraction points
  2. Trained AI to automatically classify incoming documents correctly
  3. Extracted relevant data from incoming documents relevant for the overall claims handling process
  4. Introduced automatic filing and assignment of claims files directly into underlying core claims handling system
comparison

Claims indexation

  • omni:us
  • Manual

AI-Powered Document Processing

Selected-file
DOCUMENT CLASSIFICATION

Classify documents or pages into predefined categories. Based on visual and textual information. Use of state of the art deep learning based classification methods.

Layout-top-panel-4
INFORMATION EXTRACTION

Extract information from semistructured documents using state using advanced named entity recognition and end2end information extraction techniques.

Pen-tool-vector
HANDWRITTEN TEXT RECOGNITION

Extract information from handwritten forms using language and writer independent handwritten text recognition technology.

Chat#2
FREE TEXT INTERPRETATION

Interpret unstructured documents and text with text classification, named entity recognition, language models and question answering techniques.

platform
Company
Resources
Find us
Contact

omni:us is proud to be one of Europe‘s Horizon2020 SME-Instrument and Fet Flag Champions.
We are grateful to have received funding from the European Union’s Horizon 2020 Research and Innovation Programme under Grant Agreement No 820323; 850053.

Co-financed by European Fund for Regional Development (EFRE)
Pro Fit-Project “Vollautomatisierung der Wertschöpfungskette im Digitalisierungsprozess von Archivdaten” with support of IBB/EFRE in 2016/2017.

Copyright © 2024 Qidenus Group GmbH. All rights reserved.

Privacy | Legal Disclosure