Data Scientist, Financial Crime & Fraud Analytics

Role Summary

DataSeers is looking for a Data Scientist focused on fraud detection, financial crime analytics, behavioral intelligence, anomaly detection, and transaction analysis.

Unlike many traditional data-science environments, much of DataSeers' large-scale data preparation, feature generation, relationship analysis, and analytical processing is performed using HPCC Systems and ECL .

The ideal candidate understands both statistical/data-science concepts and how to work with very large financial datasets in HPCC.

This person will help improve the intelligence behind FraudSeer, CrimeSeer, IdentitySeer, and related DataSeers products.

What You Will Do

• Analyze large volumes of financial transaction and customer behavior data using HPCC Systems and ECL .
• Develop ECL-based analytical workflows and feature-generation pipelines.
• Create behavioral baselines for customers, accounts, counterparties, devices, and transaction channels.
• Develop fraud and anomaly-detection methodologies.
• Analyze first-party, second-party, and third-party fraud.
• Identify unusual transaction velocity, amount, frequency, geography, channel, and relationship patterns.
• Analyze relationships between senders, receivers, accounts, businesses, devices, IP addresses, geographies, and payment instruments.
• Build features for account takeover detection, mule activity, transaction anomalies, duplicate payments, relationship changes, and behavioral shifts.
• Perform network and relationship-based analysis.
• Evaluate existing fraud and AML rules against historical data.
• Identify opportunities to reduce false positives without reducing detection effectiveness.
• Develop statistical and machine-learning models where appropriate.
• Work with labeled and unlabeled datasets.
• Design experiments and backtesting approaches using historical transaction data.
• Collaborate with FraudSeer, CrimeSeer, IdentitySeer, ETL, and platform engineering teams.
• Translate analytical research into production detection logic.
• Develop interpretable and explainable approaches appropriate for financial institutions.
• Research emerging financial crime and fraud typologies.

Required Experience

• Degree in computer science, statistics, mathematics, data science, engineering, economics, or another quantitative discipline.
• 3+ years of professional data science, quantitative analytics, or fraud analytics experience.
• Hands-on experience with HPCC Systems and/or ECL , or significant willingness and demonstrated ability to become productive in ECL quickly.
• Strong SQL skills.
• Strong statistical and analytical skills.
• Experience working with very large datasets.
• Understanding of classification, clustering, anomaly detection, feature engineering, statistical testing, and model evaluation.
• Ability to analyze complex relationships between entities and transactions.
• Ability to explain analytical results clearly to product, engineering, compliance, and business teams.

Preferred Experience

• Strong HPCC/ECL experience.
• Fraud detection or financial crime analytics experience.
• Banking, fintech, cards, or payments experience.
• Graph or network analytics.
• Time-series and behavioral analytics.
• Account takeover, mule detection, synthetic identity, payment fraud, or transaction monitoring experience.
• Python experience for research, experimentation, or model development.
• Experience taking analytical models from research into production.
• Familiarity with AML transaction monitoring and regulatory expectations.

Important Distinction

This is not primarily a Python notebook position.

Python, R, and traditional machine-learning frameworks may be used where appropriate, but the successful candidate must be comfortable performing substantial analytical work against large-scale financial datasets using the DataSeers HPCC/ECL environment.

What Success Looks Like

You should be able to start with billions of financial transactions, use HPCC/ECL to transform those transactions into meaningful behavioral and relationship features, identify patterns associated with fraud or financial crime, validate those patterns historically, and work with engineering to turn them into scalable production detection capabilities.