DEEP PATEL

Analyzed 5M+ healthcare and claims records in SQL and Python to identify cost, utilization, and processing-delay drivers, reducing claim processing time by 25% and improving reporting accuracy. Led migration of marketing and sales datasets from on-premise SQL Server to Microsoft Fabric using Azure Key Vault for credential management, cutting manual data-prep effort by 50%. Built scalable ETL pipelines in Microsoft Fabric with PySpark/Polars/Spark SQL using a Medallion architecture to raise data quality by 30% and reduce downstream processing time by 40%. Standardized transformations into reusable Fact and Dimension models to enable self-service and reduce redundant ad-hoc requests by 25%. Delivered Power BI and Tableau dashboards for patient flow, cost drivers, and KPIs by optimizing semantic models and incremental refresh to cut executive load times by 35%. Partnered on incremental ingestion/validation workflows aligned to compliance standards and documented lineage and transformation

Analyzed 8M+ transaction, loan, and customer-master records in SQL and Python for a retail banking and lending client, producing portfolio views of delinquency, product mix, and revenue concentration for monthly credit committee reviews. Modeled customer profitability and cross-sell propensity across deposit, card, and loan products, informing targeted campaigns that increased product revenue by 15% and improved conversion in prioritized segments. Automated 20+ Excel and BI regulatory/management reports covering exposure, collections, and daily balance movements, reducing turnaround time by 40% and eliminating manual reconciliation errors. Designed A/B tests and hypothesis tests for digital banking and collections outreach, producing statistically validated recommendations for go-to-market and retention strategy, and created SQL query libraries and data dictionaries used for analyst onboarding.