Data scientist, fraud risk mitigation

Role Summary

Data Scientist supporting fraud strategy development, fraud analytics, and operational monitoring across a consumer lending portfolio. Working under the fraud strategy lead and alongside senior team members, you will analyze fraud trends, monitor portfolio and decisioning performance, and help translate findings into policy and rule recommendations for the decision engine.

This is an individual-contributor role for someone with strong analytical fundamentals and hands-on SQL and Python skills who wants to build depth in fraud risk. Day-to-day work is weighted toward analysis, recurring reporting, monitoring, and cross-functional execution with fraud operations, product, credit/risk, and data engineering — including scorecard and model performance monitoring.

Key Responsibilities

Analyze fraud trends and emerging patterns across application, behavioral, device, and third party data, and summarize findings with clear supporting evidence. Monitor fraud performance metrics and operational outcomes — loss rates, capture rates, false positive rates, approval impact, vintage trends, and segment-level KPIs — and flag issues for review. Support policy, rule, and threshold recommendations for the decision engine, sizing expected impact and recommending changes with guidance from the fraud strategy lead and senior team members. Prepare recurring fraud reports and deep dives — loss attribution, typology trends, and decisioning outcomes — with commentary on drivers of change. Track fraud scorecard and model performance (PSI, score drift, KS, decay) and monitor reason code and segment-level behavior, escalating signs of degradation. Evaluate third-party fraud and identity signals (identity verification, device intelligence, consortium data, bank/transaction data) and contribute benchmarking analysis to onboarding, retirement, and reweighting decisions. Support test-and-learn analyses — champion/challenger tests, policy backtests, and holdouts — including data preparation, measurement, and result summaries. Partner with fraud operations on case and queue feedback, translating investigator findings into analytical follow-ups, rule ideas, and reporting improvements. Work with product, data engineering, and decisioning platform teams to implement and monitor fraud rules and data signals, validating logic and post-deployment results. Run data quality checks on fraud reporting and analysis datasets, documenting assumptions, exclusions, and known limitations. Document analyses, methodology, and recommendations so that findings are reproducible and easy for partners to review. Contribute ideas and observations to the broader fraud strategy agenda and follow shared analytical, code review, and reporting standards.

Qualifications

1–3 years of experience in fraud analytics, risk analytics, data science, credit risk, fintech or financial services analytics, or a closely related quantitative field. Strong hands-on SQL and Python skills used for real analysis, segmentation, and reporting. Experience working with large structured/tabular datasets, including cleaning, joining, and quality validation. Experience building or maintaining dashboards and recurring reporting for business partners Understanding of fraud typologies in consumer lending — identity, synthetic, first-party, and third-party fraud — or clear curiosity and willingness to learn them. Working familiarity with how scores, rules, and thresholds drive decisions, with the ability to interpret outputs and monitor performance. Ability to work effectively with guidance — taking direction on scope and method, asking good questions, and delivering carefully checked work. Clear written and verbal communication; able to explain analyses, results, and caveats to technical and non-technical stakeholders. Bachelor's degree in a quantitative field (Statistics, Economics, Mathematics, Computer Science, Engineering, or related), or equivalent practical experience.

Preferred Qualifications

Consumer lending, fintech, or financial services experience, particularly subprime or near-prime personal loans. Experience partnering with fraud operations or investigations teams on case and queue outcomes. Exposure to third-party fraud and identity data providers and to benchmarking their signals. Familiarity with decision engines or rules platforms and how fraud rules are configured, deployed, and monitored. Experience with experimentation and measurement (champion/challenger, A/B tests, backtests, holdouts). Exposure to scorecard or model performance monitoring (PSI, KS, calibration, reason-code analysis). Exposure to identity, synthetic, first-party, or third-party fraud patterns in application and account data. Familiarity with US consumer lending regulations and risk management practices.