Data Scientist ll - Digital Intelligence

The role

You will develop machine learning features, analytical methods, and production-oriented risk signals using various telemetry. You will work with complex and noisy data to improve fraud detection and identity confidence.

What you will do

• Develop machine learning features, models, and analytical methods for device, network, browser, mobile, session, and behavioral intelligence.
• Work on scoped fraud and identity risk problems.
• Build features from large-scale, high-cardinality, sparse, noisy, and platform-dependent telemetry.
• Analyze signal patterns such as spoofing, emulator behavior, automation, proxy/VPN usage, and device fragmentation.
• Design and execute validation analyses including train/test splits, leakage review, and drift assessment.
• Use supervised, unsupervised, statistical, and heuristic approaches to identify fraud and identity risk signals.
• Investigate imperfect labels, delayed outcomes, and changing fraud patterns.
• Partner with engineering, product, risk, and platform teams.
• Contribute to model documentation, feature definitions, and dashboards.
• Support junior data scientists through code review and analytical feedback.

Required qualifications

• Bachelor’s, Master’s, or Ph.D. in Computer Science, Machine Learning, Statistics, Mathematics, Data Science, or a related quantitative field, or equivalent practical experience.
• 5+ years of experience in data science, applied machine learning, statistical modeling, analytics engineering, or a related technical role.
• Experience building, evaluating, and improving machine learning models, features, analytical pipelines, or risk signals.
• Strong SQL skills and experience working with large-scale, complex datasets.
• Strong proficiency in Python and experience with data science libraries such as pandas, NumPy, scikit-learn, XGBoost, TensorFlow, PyTorch, or similar.
• Experience with distributed data processing tools such as Spark, PySpark, Databricks, or equivalent frameworks.
• Solid understanding of supervised learning, unsupervised learning, feature engineering, model evaluation, statistical validation, and experiment analysis.
• Ability to work with noisy data, imperfect labels, missing values, instrumentation gaps, and changing data distributions.
• Strong analytical judgment across data quality, feature design, model selection, explainability, and business impact.
• Experience collaborating with engineering, product, analytics, or risk teams to move data science work toward production or operational use.
• Clear communication skills to explain technical work to non-specialist stakeholders.
• Ability to operate independently on defined problem areas.

Nice to have

• Background in fraud detection, identity verification, trust and safety, anomaly detection, cybersecurity, risk modeling, or another adversarial data domain.
• Experience with device intelligence, browser/mobile fingerprinting, behavioral biometrics, network intelligence, VPN/proxy detection, or telemetry signal processing.
• Experience developing features from high-cardinality categorical data using techniques such as aggregation, frequency encoding, target encoding, embeddings, graph features, or representation learning.
• Familiarity with production ML workflows, model monitoring, feature monitoring, or batch and near-real-time decisioning systems.
• Experience with dashboarding, model explainability, feature documentation, or customer-impact analysis.
• Interest in adversarial behavior, fraud patterns, telemetry quality, and applied ML systems that operate in real-world production environments.

About the company

Socure is building the identity trust infrastructure for the digital economy. The company uses AI and machine learning to provide digital identity verification and fraud prevention solutions.