Today, most forward-looking enterprises are adapting to the post-COVID 19 world and prioritizing AI-powered automation to transform their businesses and compete for customer satisfaction. The insurance industry is at the forefront of this transformation and among the early adapters. Insurance has always been heavily reliant on paper-based, manual processes. However, one category that stands out is the ACORD.
Most insurers operate in a legacy model, relying on third-party vendors or operational teams for manual data enrichment. The insurance business depends on the accurate processing of forms and documents, and there are significant pain points in processing complex insurance documents:
- Inconsistent document types: Agents send documents of different styles and formats, resulting in considerable data extraction.
- High turnaround time: Entering data manually from forms is time-consuming, error-prone, and expensive
- Wasted productivity: Underwriters and assistants spend up to 45% of their time on enriching data rather than on evaluating risk
- Adhering to regulation and compliance: Constant changes to insurance regulations necessitates changes to processes.
- Lack of measures on data quality, completeness, or accuracy of the data extracted by the agents manually
Most of the property and casualty insurance carriers in the U.S. use ACORD forms and certificates. These are standardized forms, governed by a nonprofit organization - the Association for Cooperative Operations Research and Development, that develops standards and process guidelines for the insurance industry. Several administrative insurance processes involve handling these forms to extract customer inputs.
Processing the ACORD forms
Processing ACORD forms is a painstakingly slow and manual activity. Insurers receive ACORD documents and supporting supplemental forms via email, fax, or web portals. Personnel then manually type the data from the forms into some downstream application. Although most ACORD forms are completed electronically, some can also be filled by hand or have checkboxes marked in pen.
Some insurers have attempted to use traditional OCR technology only to find that the legacy technology creates more problems than it solves. You could also automate the processing of ACORD forms using a template approach since the forms are standardized and structured. However, many of those approaches require labeling the documents and marking the extraction entities. Considering the variations and types of ACORD currently available, this would be a time-consuming and cumbersome task.
Another problem with this approach is that these forms come with supporting documents that may or may not be in a structured format. For example, home insurance applications come with loss run reports and statement of values; auto insurance claims come with photos, estimates from auto body companies. None of these documents are structured, so a template based automation approach or RPA tool would be out of scope.
The key to the successful transformation of the document intake process is choosing a solution that can identify the context behind documents and images without constant manual intervention.
Intelligent Document Processing for ACORD forms
A better approach is to adopt intelligent document processing (IDP) solutions to read and extract data from semi-structured or unstructured documents. IDP involves using AI technologies like natural language processing and machine learning to read forms and documents much like a human would.
Powered by its proprietary multi-stage engine, that combines OCR with computer vision and natural language processing, Kanverse‘s Insurance Document Processing solution achieves the goal of ACORD automation. Kanverse automates the ingestion of ACORD and supplemental forms across multiple channels, classifies the types and variations of the documents, performs data extraction with industry-leading 99.5% accuracy, validates the data based on business rules and files it in Insurance software systems.
Contact Kanverse.ai for more details.
About the Author
Kingshuk Ghosh, Product Manager (Kanverse.ai)