Sr Data Engineer

- Built Microsoft Fabric medallion lakehouses (bronze, silver, gold) using OneLake, PySpark notebooks, Data Factory pipelines, and Warehouse T-SQL; owned silver-to-gold transforms for 20-40 EHR, claims, and lab feeds supporting 100,000+ CKD/ESKD members. - Automated Fabric workspace, lakehouse, and warehouse promotion with Git and Azure DevOps CI/CD, cutting partner environment setup from about 2 days of manual tickets to under 30 minutes. - Built HIPAA and HITRUST PHI controls with security and compliance teams, including Microsoft Purview sensitivity labels, workspace RBAC, column-level security on 100+ sensitive fields, audit logging, and isolated dev/staging/production environments. - Built 15+ gold-layer marts in Spark SQL and Warehouse T-SQL with incremental loads, schema contracts, and pipeline CI, turning partner feeds into risk stratification, hospitalization, and total-cost-of-care datasets. - Monitored data quality rules and T+8 hour freshness SLAs (about 99% hit rate) for owned pipelines, catching late or broken clinical files before they reached care team workflows. - Built Power BI Direct Lake semantic models and Fabric Copilot reporting with analytics and clinical operations stakeholders (25-40 people), reducing recurring ad-hoc SQL requests by about 30-40%. - Tuned Fabric Spark jobs and warehouse capacity, partitioning, and incremental refresh with the platform team, contributing to a 20-30% reduction in compute spend, and supported lakehouse on-call with typical resolution time under 45 minutes. - Partnered with product, clinical, and care team stakeholders to turn data needs into data contracts and SLAs, shortening new-partner dataset onboarding from weeks to days. - Built Warehouse and Power BI semantic models with row-level security for 10+ payor and health-system partners, so each partner could see only their own members, supporting value-based care reporting without cross-tenant data exposure.

- Built Databricks medallion pipelines (bronze, silver, gold) using PySpark, Delta Lake MERGE/CDC, Auto Loader, and Databricks Workflows; owned silver-to-gold transforms for 10-20 learning, skills, performance, and HRIS/ATS feeds. - Implemented Unity Catalog schemas and access grants with the platform team through CI/CD using Databricks Asset Bundles, cutting catalog and setup tickets from about 1-2 days to under 30 minutes. - Built 25-40 production dbt models in SQL and Python with incremental logic, automated tests, and CI on pull requests, turning LMS, skills, and recruiting events into gold marts for course completion, skills-gap, and compliance-training reporting. - Monitored owned jobs with dbt tests, daily freshness checks, and pipeline alerts, maintaining 98%+ on-time delivery and catching late or broken HR and learning files before they reached People Analytics dashboards. - Built Databricks Genie reporting on Unity Catalog governed gold tables with analytics and HR partners (15-25 stakeholders), reducing recurring ad-hoc SQL requests by about 25-35%. - Built Algolia indexing jobs on curated course, skills, and people records with product partners, cutting employee search response time to under 200 milliseconds. - Tuned Databricks job clusters and Delta tables using Photon and partitioning/liquid clustering with senior engineers, contributing to a 15-25% reduction in DBU spend. - Shared lakehouse on-call rotation, recovering failed learning and HR jobs with typical resolution time under 60 minutes, and documented runbooks for the team. - Built bronze-layer ingestion from Cornerstone REST/OData and HRIS/ATS APIs, landing incremental employee and course data so L&D reporting stayed current across 50+ customer tenants.