What You Will Do
Design, develop, and evaluate machine learning models for fraud detection, including fraud scoring and ensemble models that combine multiple sources of evidence.
Analyze large-scale datasets to identify complex patterns, relationships, behaviors, and anomalies, developing new signals to improve fraud detection.
Define robust experimentation and evaluation methodologies, including model calibration, operating thresholds, and performance metrics under realistic production conditions.
Develop interpretable and explainable ML approaches that help internal stakeholders and customers understand fraud scores and decisions.
Explore innovative approaches to fraud decisioning, including human-in-the-loop techniques and causal methods.
Collaborate with Product and Fraud Domain Experts to translate real-world fraud challenges and production feedback into effective machine learning capabilities.
Partner with Data Engineering to transform ML prototypes into scalable, maintainable, production-ready solutions.
- Contribute to model monitoring, governance, technical documentation, ML tooling, and shared best practices across the team.
- What You Need
- A Bachelor's, Master's, or PhD degree in Computer Science, Statistics, Mathematics, Machine Learning, or a related quantitative field, or equivalent practical experience.
Skills and expertise typically gained through approximately 5 years of experience as a Data Scientist, Applied ML Scientist, or in a comparable role.
Strong knowledge of statistical methods, supervised machine learning, and model evaluation, particularly for highly imbalanced classification problems.
Hands-on experience in feature engineering and machine learning model development using complex behavioral, relational, or contextual data.
Strong programming and analytical skills using Python, PySpark, and SQL.
- Experience working with cloud-based machine learning and data infrastructure.
