DescriptionThe Opportunity
Reporting to the Manager, Enterprise Fraud, the Senior Data Specialist is a core member of the Enterprise Fraud Analytics team. This hybrid role bridges the gap between data engineering and data science, focusing on architecting scalable data pipelines and designing advanced AI/ML models to detect and prevent insurance fraud. The candidate will also support Enterprise Fraud strategic priorities and support Fraud Savings initiatives.
Operating in a highly collaborative environment, this role works alongside Business Analysts and the Special Investigations Unit (SIU). The Senior Data Specialist will heavily leverage Google Cloud Platform (GCP), BigQuery, Vertex AI, and Gemini. This role is critical in transforming complex fraud detection requirements into robust datasets and extracting predictive signals from unstructured text and claims data, ensuring both architectural scalability and mathematical rigor.
What to Expect
-
Data Architecture & Pipeline Engineering
Design, build, and maintain highly scalable ELT/ETL data pipelines using modern cloud infrastructure (GCP, BigQuery, GCS). Architect processes to ingest, transform, and integrate large volumes of structured claims data and unstructured third-party data to build the foundational datasets required for advanced fraud analytics.
-
Algorithm Design & AI Model Development
Design, train, and iterate on predictive machine learning models and AI solutions. Focus heavily on Natural Language Processing (NLP) and Generative AI (e.g., prompt engineering with Gemini) to extract actionable fraud indicators from unstructured SIU data (case notes, medical records) and structured policy data.
-
Model Evaluation & Explainability
Conduct rigorous offline model evaluation (precision, recall, slice analysis). Champion Explainable AI (XAI) by utilizing methods like SHAP values or LLM reasoning traces to ensure model outputs are transparent, interpretable, and trusted by non-technical SIU investigators.
-
Exploratory Data Analysis (EDA) & Automation
Work closely with SIU stakeholders to understand emerging fraud schemes. Conduct in-depth EDA using SQL and Python to assess data feasibility and establish baseline metrics. Design and maintain internal automations and scheduled d