data
Posted Feb 23Senior Data Scientist - Fraud Prevention
at Neighbor
United StatesRemote
Responsibilities
- Design and Lead Key Analyses and Metric Evolution
- Analyze large, complex datasets to identify abuse patterns, fraud signals, and harmful behavior trends.
- Conduct root cause analysis to diagnose safety incidents and emerging risks.
- Evaluate new tool effectiveness (including AI), and impact on agent efficiency and user satisfaction.
- Define and track core metrics (e.g., harm prevalence, violation rates, detection accuracy).
- Build dashboards and reporting frameworks to track platform health and safety performance.
- Develop Models & Rules
- Develop heuristics, statistical models, and machine learning solutions for proactive detection of abuse, fraud, or harmful content
- Build prediction systems (e.g., anomaly detection, risk scoring, behavioral profiling).
- Improve automated enforcement and moderation workflows.
- Evaluate model performance and iterate on detection strategies.
- Evaluate Product / Policy Changes Via Experimentation
- Design and analyze experiments (A/B tests, causal inference) to measure safety feature impact (e.g., login & verification, AI moderation support).
- Quantify tradeoffs between operational efficiency, safety, and user growth/experience. Guide TnS team on key tradeoffs in decision-making
- Own Cross-Functional Partnership
- Influence decision-making through data storytelling and insights.
- Standardize analytical methodologies and tools for scalable decision-making.
Requirements
- experience by developing strong data foundations, robust metrics and measurement frameworks, and advanced analytics and modeling that improve harm prevention and detection and moderation effectiveness.
- Bachelor’s or Master’s degree in Statistics, Computer Science, Mathematics, Economics, or a related quantitative field.
- experience working with large-scale data and statistical analysis, including 1+ year of data science
- experience in fraud prevention, moderation, or risk.
- Proficiency in SQL and at least one scripting language (e.g., Python or R ).