We have developed successful predictive modeling solutions for fraud detection and compliance management programs in many sectors, including:
Tax & Revenue
Utilize a variety of predictive models to screen tax returns for fraudulent refund requests, identity theft, and other fraud schemes. We can also identify candidates for audit and prioritize non-filer cases based on tax potential.
Enabling online merchants detect fraudulent credit card use for online purchases.
Scoring models enable insurance companies to identify inflated and false claims, lowering the cost related to fraud and abuse.
Using anomaly detection models, we helped OEM manufactures identify repair facilities that submit false or misrepresented claims.
Using anomaly detection techniques, we helped government agencies identify retail stores that trafficked cash for food assistance credits.
Our predictive analytics and anomaly detection solutions for fraud leverage an experiential foundation that spans many industries. Our solution patterns reutilize effective schemes we’ve found to be applicable across many domain areas. These design patterns leverage two fundamental aspects of fraud:
- Fraudulent events, or transactions, typically following subtle but repeatable patterns
- Fraudulent transactions often appear as anomalies compared to legitimate transactions, when compared across one or more dimensions.