Turning messy project data into decisions.

MEK Operations & Analytics

As Director of Operations, I run the analytics, forecasting, and project management behind MEK's active portfolio for company leadership — building the revenue models, labor-margin tracking, and executive dashboards, while directing full project lifecycles across a multi-site public works and school-district portfolio spanning San Diego and San Francisco.

The Analytics, In Practice

From crew budgets to executive insight — and the tools that keep it running.

Mock data for illustration — site names and chart values are illustrative, not MEK's actual numbers. The 18% first-year labor cost reduction is a real MEK result.

CMS Medicare Provider Utilization & Cost Analysis

Health plan analysts need to know where medical spending actually concentrates — by procedure, provider, care setting, and specialty — to catch cost drivers and manage risk. I used CMS's public Medicare Physician & Other Practitioners dataset as a real-world proxy for claims data, and built the kind of analysis a payer's data team runs day to day.

Designed a PostgreSQL schema from CMS's 2024 public release, filtered to California (832,561 provider/procedure/place-of-service rows), and wrote five SQL queries of increasing complexity — ranked aggregations, a CASE-based relabeling with a window function for care-setting spend share, and a specialty-level rollup pairing total spend against provider count. Turned the results into narrative chart visuals with Python/matplotlib, all version-controlled on GitHub.

The Findings

$12.8B in California Medicare spend, broken down by setting, specialty, and provider.

Real, public CMS data (2024 Medicare Physician & Other Practitioners release, California) — analysis and figures above are mine, not MEK's. Parts 2 (pharmacy claims ETL & data quality validation) and 3 (claims-based readmission forecasting) of this sprint are still in progress.