Responsibilities
- Design and implement machine learning features, models, and analytical techniques focused on device, network, browser, mobile, session, and behavioral data.
- Address defined fraud and identity risk challenges requiring careful evaluation of data quality, labeling accuracy, telemetry coverage, and product tradeoffs.
- Extract meaningful features from large, high-cardinality, sparse, and noisy telemetry data influenced by platform-specific behaviors.
- Detect and analyze patterns indicative of spoofing, emulator use, automation tools, proxy or VPN connections, low-entropy fingerprints, missing telemetry, and fragmented device or session data.
- Conduct validation analyses including train/test partitioning, holdout testing, leakage detection, concept drift monitoring, customer impact assessment, and feature stability checks.
- Apply supervised and unsupervised learning, statistical modeling, and heuristic methods to uncover persistent fraud and identity risk indicators.
- Examine issues related to unreliable labels, delayed outcomes, gaps in instrumentation, and evolving fraud tactics to separate true signals from data noise.
- Collaborate with senior data scientists, engineers, product managers, risk analysts, and platform teams to define requirements, process data, build features, and enable production deployment.
- Support the creation of model documentation, feature specifications, explainability reports, monitoring dashboards, and production readiness evaluations.
- Clearly communicate modeling approaches, underlying assumptions, key findings, limitations, and strategic recommendations to both technical and non-technical audiences.
- Mentor junior data scientists and analysts through code reviews, analytical guidance, and knowledge sharing on modeling and validation best practices.
