Data Scientist Predictive Utilization Analytics

Job Description Must Have Technical/Functional Skills

Data Scientist to build a Predictive Benefit Utilization Model and an Attribution Model leveraging Microsoft Fabric, with the objective of improving accuracy over current manual / Excel‑based projections and enabling data‑driven executive decision‑making

Roles & Responsibilities

• Design, develop, and validate a Baseline Utilization Prediction Model using historical claims and member data (demographics, geography, plan design, diagnosis/procedure codes)
• Develop an Attribution / Explainability layer to identify key drivers of deviation between expected and actual utilization (where data permits)
• Perform feature engineering on structured insurance datasets sourced from on prem SQL DW and curated in Microsoft Fabric / OneLake
• Apply appropriate statistical, machine learning, or time series techniques to improve forecast accuracy versus current baseline methods
• Conduct model validation and back testing, comparing predicted utilization against actuals and existing manual projections
• Ensure model explainability (feature importance, drivers, narratives) suitable for executive and business stakeholder consumption
• Collaborate with Data Engineering to ensure data readiness, quality checks, and semantic alignment in Fabric
• Support Power BI / Fabric semantic model consumption, ensuring model outputs can be operationalized in reports and analytics layers
• Document modeling approach, assumptions, limitations, and outcomes for stakeholder review and next phase decisioning

Required Skills & Experience Core Data Science

• Strong experience in predictive modeling, regression, classification, and/or time series forecasting
• Proficiency in Python (pandas, numpy, scikit learn, statsmodels or equivalent)
• Experience with model validation techniques, accuracy metrics, and performance comparison

Domain Experience

• Prior experience with insurance, healthcare, or benefits utilization data (claims, eligibility, plan design preferred)
• Ability to interpret and translate business drivers behind utilization patterns

Platform & Tools

• Experience working in cloud analytics platforms (Microsoft Fabric preferred; Azure / Databricks acceptable)
• Familiarity with SQL based data sources and collaboration with ETL / data engineering teams
• Exposure to Power BI or semantic models is a plus (not mandatory)
• Experience with explainable AI (XAI) techniques
• Prior exposure to Microsoft Fabric Copilot / AI assisted analytics concepts

Salary Range: $90,000 to $115,000 per year

Qualifications: BACHELOR OF COMPUTER SCIENCE