Role summary
FWA Data Science role is responsible for transforming validated analytical prototypes into robust, scalable, and measurable detection models ready for production implementation. The role owns model enhancement, feature engineering, calibration, validation, explainability, performance assessment, and continuous model improvement throughout the solution lifecycle. Working closely with Actuarial, Claims Integrity, and FWA AI Engineering teams, the role ensures detection models remain effective, relevant, and aligned with evolving fraud patterns after deployment.
Responsibilities
- Assess AI/ML prototypes received from Actuarial teams
- Refine model logic and improve statistical rigor
- Enhance feature engineering methodologies
- Validate and clarify prototype assumptions
- Develop explainable detection methodologies
- Support migration of approved analytical models into cloud-based solutions
- Improve and retrain FWA models using new data and investigation outcomes
- Develop advanced fraud detection features
- Enhance anomaly detection models
- Build risk scoring methodologies
- Improve model precision and operational effectiveness
- Define model success metrics
- Monitor precision, recall, and business impact
- Perform false-positive and false-negative analysis after go-live
- Validate model outputs with investigators
- Support production readiness reviews
- Analyze investigation outcomes
- Identify emerging fraud behaviors
- Design new detection features
- Recommend modifications to existing models
- Collaborate with FWA operations on evolving threats
- Monitor business effectiveness of deployed models
- Recommend retraining and recalibration activities
- Support periodic model reviews
- Maintain model documentation and governance artifacts
Education
Bachelor's degree
Education · Experience · Type
Bachelor'sSalary not stated