Case Studies

Claim Reconciliation Agent

Claim Reconciliation Agent

Artificial Intelligence

Challenge

Navigating millions of disparate medical documents presents a substantial bottleneck. Manual data extraction and format inconsistencies increase the risk of errors in carrying over such repetitive tasks. This complexity prolongs reconciliation times execution and exposes the firm to higher compliance risk.

Solution

A digital platform to assist users that process claims during the reconciliation process. The platform integrates document ingestion pipelines, data extraction and user interfaces. The ingestion pipeline classifies a document in its category and extracts the relevant information set. Every piece of data is linked to the source where it is contained for ease of validation from the user interfaces.

Benefit

More than 5 millions documents managed over the past years. Every document is processed in less than 5 seconds. On average, 85% of the required information from a document is extracted automatically and the remaining 15% requires human intervention either for retrieval or for multi answers disambiguation. The overall result is a process time reduced by 55%.

Quote

With this application the liquidation process has been speeded up by 50%. Allowing to improve the adherence to the liquidation SLAs.