15% reduction in premium using ML-backed underwriting system
Built an ML-backed underwriting system for premium reduction and assign an overall group risk score using lab/diagnostic, prescription, and claims data at multiple levels and data bridges
Business Challenges
The client is a leading healthcare AI company catering to a multitude of health insurance providers and pharma companies. They were looking for a data and ML-backed underwriting system to minimize the cost of claims and offer a competitive premium while ensuring faster turn-around time for the claim processing. The existing system was based on manual rules called expert systems and lacked any predictive model.
Sigmoid Solution
Sigmoid bridged all the datasets to extend the de-identified patients’ information to cover a multitude of features across claims, lab and prescription data. We built machine learning models to generate a group risk score based on individual patient statuses for each employee group. We also provided recommendations regarding bidding costs, improving existing pricing and opt-out deals while creating a framework for automated report generation in turn ensuring faster turnaround time of claim processing and reduced risk of uncertainty.
Business Impact
With 50% improvement in turnaround time, our ML-backed underwriting system led to 15% reduction in cost of premium and 8% increase in margins.
Relevant Case Studies
ML Model Improvement & Management using MLOps
ML-based demand forecasting solution to improve forecast accuracy while reducing costs by 5x
