John Hancock is seeking to solve the business problem of fraud detection and risk mitigation within their long term care insurance business by developing cutting-edge models and innovative solutions to protect the organization and policyholders from fraudulent activities.
Requirements
- Advanced proficiency in Python or R for statistical analysis and machine learning
- Expert-level SQL skills and experience with database management systems
- Hands-on experience with machine learning frameworks
- Proficiency with graph analytical methods and libraries
- Experience with time series methods and libraries
- Demonstrated expertise in time series forecasting, trend analysis, seasonality detection, and change point detection
- Strong background in outlier detection methodologies including statistical approaches, machine learning methods and deep learning techniques
Responsibilities
- Design and build sophisticated fraud detection models with emphasis on time series analysis to identify temporal patterns and trends in fraudulent behavior
- Develop anomaly detection systems to flag unusual claims patterns, provider behaviors, and policyholder activities
- Create graph-based models to uncover fraud rings, provider networks, and suspicious relationship patterns
- Build ensemble models that combine temporal, network, and statistical approaches for comprehensive fraud detection
- Perform advanced statistical analysis on large, complex datasets to uncover fraud indicators
- Design and implement digital controls and automated workflows to mitigate fraud impact
- Develop innovative analytical solutions to address emerging fraud schemes and attack vectors
Other
- 2-4 years of experience in data science, analytics, or machine learning roles
- Master's degree in Statistics, Mathematics, Physics, Engineering, Computer Science, or other quantitative science discipline
- Ability to partner with claims, underwriting, and compliance teams to implement analytical solutions
- Ability to present findings and recommendations to senior leadership and cross-functional teams
- Ability to document methodologies, model logic, and analytical processes for regulatory compliance
