CMS Medicare Part D Analytics

Build a Medicare Part D data pipeline using DuckDB by ingesting SPUF 2025 Q3 data into a medallion architecture, transitioning raw Bronze records into curated Gold and ML layers for fast, auditable analytics.

Compare Medicare Part D plan formulary costs across counties by applying geospatial and affordability modeling to SPUF 2025 Q3 data, generating county-level metrics and pre-computed plan rankings for beneficiaries.

You can use machine learning to generate Medicare Part D recommendations by processing curated Gold layer data through an ML model layer to produce top plan rankings for synthetic beneficiaries.

This DuckDB medallion architecture supports network access and distance proxy modeling by transforming raw CMS Part D data through pipeline scripts that compute geographic access metrics for county-level plan comparisons.

Rank cost-effective Medicare Part D coverage for insulin users by running the data pipeline to generate formulary metrics, distance proxies, and pre-computed recommendations tailored to specific beneficiary scenarios.