Hybrid Senior Data Scientist – Fraud/Risk

at Fiserv

• Job title: Senior Data Scientist – Fraud/Risk
• Job type: Full Time
• Experience level: Senior
• Salary: $111,000 - $188,400 per year
• Degree requirement: Bachelor's Degree
• Tech skills: Azure, Kafka, Neo4j, PySpark, Python, Scikit-Learn, SQL
• Location requirements: Hybrid, Alpharetta, United States

About the role

• Senior Data Scientist developing AI-driven fraud detection solutions at Fiserv. Architecting products to analyze massive transaction streams for risk signals.

Responsibilities

• Monetize Risk Intelligence: Architect and deploy production-grade AI/ML frameworks to monetize Fiserv’s unique data footprint into inventive risk scores and insights that detect identity, transaction, and business-level threats.
• Architect Financial AI: Build custom GenAI, NLP, and LLM models for high-velocity stream processing, focusing on extracting risk indicators and behavioral anomalies from structured transaction data and unstructured metadata.
• Next-Gen Frameworks: Implement LangChain and LlamaIndex to develop RAG and Agentic AI frameworks that enable institutional clients to query and interact with complex, multi-dimensional risk datasets.
• Quantitative Collaboration: Work in a high-performance team environment, collaborating with Product Managers, payment system experts, and Engineering to deploy and monitor production-grade AI and ML models.
• Strategic Synthesis: Distill complex quantitative risk insights into high-level investment and risk theses for executive leadership and sophisticated external stakeholders.
• Data Stewardship & Compliance: Partner with the Data Usage Committee, Model Governance, Legal, and Compliance teams to ensure data privacy and adherence to strict data usage rights within the DCS framework.

Requirements

• 7+ years of experience leveraging large scale datasets to develop tactical insights into fraud typologies such as ATO, Synthetic ID, and AML using ML, RAG, and NLP.
• 7+ years of experience formulating research problems, designing champion/challenger back-tests, and implementing production-ready solutions in a financial or high-growth tech environment.
• Experience with anomaly detection, credit risk modeling, and adversarial machine learning within merchant and banking ecosystems.
• Mastery of high-precision classification, anomaly detection, and clustering techniques, focusing on non-stationary time series analysis, Bayesian inference, causal analysis, and survival analysis to model risk probabilities, event timing, and evolving fraud trends.
• Expert proficiency in Python, SQL, and PySpark for high-volume transaction processing, with hands on use of Scikit-learn, XGBoost, LightGBM, and Deep Learning and Agentic AI frameworks for threat hunting, and graph databases like Neo4j or Tiger graph for fraud network analysis.
• Experience with Databricks and Snowflake, SageMaker or Azure ML, feature stores (e.g., Tecton, Feast) and streaming architectures (Kafka, Flink).
• Proficiency in tokenization and embeddings, with hands-on experience tuning and deploying Large Language Model architectures such as LLaMA, BERT, or Transformers.
• Bachelor’s degree in a quantitative field such as Computer Science, Mathematics, Artificial Intelligence, Financial Engineering, or Statistics.

Benefits

• Annual incentive opportunity, which may be delivered as a mix of cash bonus and equity awards