Integrating AI automation and triage portal for unstructured data management with a focus on accuracy and efficiency
Case Summary
The Challenge
The client manually processed incoming data from customers and other sources, often in multiple file types that included images. Extracting data from these files was difficult and time-consuming. Simpler automation solutions were insufficient due to the presence of non-standard data, such as footnotes and exceptions.
The insurer sought a fast, efficient, and accurate data extraction and reconciliation system that would eliminate the need for manual labor.
The Solution
SLK leveraged our Quality Center of Excellence (QCoE) and an integrated AI system to provide a simple solution for data processing. We created a triage portal where the AI would input extracted data in a predefined, standardized format.
To overcome errors caused by unstructured and/or non-standard data in the documents, a human-in-the-loop (HITL) process was implemented to review and correct any data input. This resulted in highly accurate data flows to business workflows, improved governance, and significantly reduced human dependencies.
Business Impact
SLK established 4+ CoEs to cater to multiple areas of business and provided a holistic solution for data management needs of the insurer.
>95%
Accuracy achieved in data processing
40%
New submission and endorsement processes automated
8X
Improvement in efficiency