Machine Learning with Azure Data

Day-to-Day Responsibilities

• ML Pipeline Development & Maintenance – Builds and maintains PySpark-based training pipelines for 30-day and 90-day CGM adherence prediction models using Azure Databricks and MLflow for experiment tracking and model versioning.
• Model Retraining & Validation – Executes end-to-end model retraining cycles including out-of-time (OOT) validation, risk segment analysis, and threshold optimization (Youden's J).
• Feature Drift Monitoring – Monitors production feature health using PSI (Population Stability Index) calculations and SQL-based drift detection across Marketo, PIMS, and Genesys data sources.
• Data Pipeline Engineering – Develops and maintains Azure Data Factory (ADF) pipelines for data ingestion, backfill orchestration, and scoring workflows.
• Bug Investigation & RCA – Diagnoses and resolves critical production issues (data normalization bugs, race conditions, date logic errors) with formal root cause analysis documentation.
• Documentation & Reporting – Produces deployment guides, user stories (Azure DevOps), RCA reports, and management-facing PowerPoint summaries.