I turn messy, real-world data into clear, decision-ready answers.
Pipelines, analysis, and dashboards that people actually use — built end to end.
Metals Fabrication Inc.
Airway Heights, WA
Automated cross-department VBA reporting workflows for the Estimating & Purchasing teams, eliminating an estimated 150+ hours of manual reporting a year.
Built an end-to-end document digitization pipeline in Python & OpenCV, converting ~3,000 legacy printed spreadsheets into structured, analysis-ready data.
Built an OpenAI-API lead-generation agent that turned vague, non-technical project descriptions into structured research output (GC, funding source, scope, estimated value).
Designed a Power BI dashboard tracking employee retention — trusted to work directly with sensitive HR data.
Project Tracking Dashboard
An internal web app I built to run 10+ concurrent residential remodels — replacing a fragmented mix of calendars, sticky notes, and Floorzap. Per-project task lists, status stages, and lead-source tracking, used daily across every active job.
X-Vector Speaker Embedding Systems
Undergraduate thesis on speaker representation. Built an audio preprocessing and Mel-scale filter-bank feature pipeline in Python, paired with a detailed architectural analysis of TDNN-based x-vector recognition systems following Snyder et al. (2018).
Hospital Quality & Readmission Analysis
Benchmarking a hospital network against regional and national CMS quality measures. A re-runnable ETL pipeline over four CMS Open Data APIs feeds an analysis of Sacramento-region readmission and patient-satisfaction trends, surfaced through an interactive scorecard dashboard.