A P&C insurer faced difficulties in reducing the time, resources, and cost of its quality assurance program.
Digital Intelligence Solution
EXL partnered with the client to help it move from auditing 100% of a population to a smaller sample size. To do so, EXL used predictive modeling techniques to split the quality insurance workflow into two separate queues. Transactions with a low probability for errors went into a queue that bypassed the audit phase, while those with a high probability were audited. This allocation was performed using data treatments for outliers and missing values, as well as logistic modeling techniques developed using selected variables and stratified regression across multiple lines of business.
- 25% productivity improvement
- 99% accuracy
- $300K saved over three years