At Anduril Industries, I cleaned and modeled Quality & Manufacturing data into a governed, model-ready foundation on a Spark lakehouse. I also built a QMS knowledge base and verified-procedure Q&A agent over curated data, with an estimated ~$1.2M/yr productivity value across a ~300-person organization.
At Epirus, Inc., I modeled Oracle Fusion PO revision history into the Cost Delta Tracker and developed an 11-category root-cause classification for variance analysis. The work identified a $2.3M/yr cost-leakage pattern.
At Yuji Lighting (USA), I designed analytics pipelines across a five-phase RFQ/RFP process and built Power BI dashboards that cut order-to-ship time 18%. I also unified Pipedrive and ShipStation data with Python ETL and shipped Python and Rust PDF-extraction tooling adopted as a shared library.
My data work spans defense, manufacturing, supply chain, and energy. Earlier, as a U.S. Navy submarine Sonar Technician, I worked with real-time streaming sensor data and applied data-quality and anomaly-detection patterns to track next-generation foreign submarines.
