CLIENT OVERVIEW
The specialty insurer is part of a larger property and casualty mutual insurance company based in the Midwest. It writes multiple lines of business, including professional liability, environmental, property, financial, and excess casualty, and continuously rolls out new lines of business to support customers.
BUSINESS CHALLENGE
The client wanted to modernize its claims processes, automating where possible and reducing manual workflows. The insurer knew that this investment would not only improve data accuracy and enhance operational efficiency, but also reduce loss adjustment expense (LAE) and help the organization transform claims into a strategic lever for continued growth.
Manual claims data extraction from emails and attachments was time-consuming, not to mention prone to human error and inconsistencies. These documents and emails were not in a standard and consistent format as every TPA sent the information in a different format.
There was also a lack of quality control with the existing system; the automation in place extracted data directly from emails and input it as First Notice of Loss (FNOL) without a standardized data validation process. Manual classification processes were also leading to misclassification across claim types, along with inconsistent data handling and policy validation across claims.
SOLUTION OFFERED
Having worked with ValueMomentum previously as a strategic partner on complex modernization and data projects, the insurer knew that ValueMomentum had the claims domain knowledge and the technological prowess to guide the move from manual processes to AI-driven claims workflows. ValueMomentum helped the insurer define an intelligent claims workflow approach to infuse AI-powered data extraction into its claims processes.
A focal point of the effort was using intelligent document processing (IDP) to extract both unstructured and structured data of all file formats from emails and attachments, no matter which TPA the submission came from, then conduct intelligent field mapping with mandatory claim type classification. Using a hybrid AI strategy, the project implemented context-aware structured data extraction with a dynamic metadata tagging and contextual enrichment protocol. The data was then ingested into a standard FNOL structure that was used for FNOL creation directly into Guidewire ClaimCenter.
The IDP solution was integrated with the insurer’s specialty portal to enable a systematic review process with human-in-the-loop validation prior to FNOL creation. In addition, ValueMomentum ensured that the structured data format would be consistent across all claims to avoid misclassifications and reduce the margin for error. Mandatory claim categorization (e.g., cybercrime, professional services, etc.) with automated date and state code conversion and verification with proper formatting was implemented to ensure that all claims data and FNOL creation followed a taxonomy standardized by claim type.
The project has helped the insurer not only improve the efficiency and accuracy of its claims management processes, but has also laid the groundwork for the company to incorporate AI-based workflows across the organization.
VALUE DELIVERED
Infusing IDP and automation across the claims management life cycle helped the insurer cut its manual data entry time by 80% and reduce its LAE. It also created a more consistent process and format for extracting structured data from complex documents and streamlined the workflow from email receipt to FNOL creation.
Mandating a human review step within the automated process helped the carrier ensure accuracy prior to FNOL submission, while standardizing claims type classification across all specialty lines has helped reduce misclassification. This categorization has also delivered better overall risk management for the insurer and reduced processing delays through automated extraction and validation.
The project has helped the insurer not only improve the efficiency and accuracy of its claims management processes, but has also laid the groundwork for the company to incorporate AI-based workflows across the organization. The lessons learned from this effort provided frameworks and governance practices that can be leveraged for future AI-fueled initiatives across claims and the broader enterprise.
