Ph.D. applied scientist and engineering manager building production machine learning, statistical evaluation, and data platforms in complex, regulated environments.
I lead Business Intelligence & Operations Surveillance at the Washington State Department of Corrections (Research & Data Analytics). Recent work includes a production multi-label NLP system over 110K+ records and 73 labels (scikit-learn / TF-IDF; 75.6% micro-F1, 88.3% precision; k-fold and Jaccard evaluation; human-in-the-loop iteration), SQL ETL (T-SQL, Oracle) feeding Power BI at scale, automation that cut reporting cycle time roughly 60-70%, and experiment design (A/B and quasi-experiments; chi-square, Cramer’s V) on large administrative and privacy-constrained healthcare claims cohorts – including geospatial and network analytics (GeoPandas, NetworkX, Quarto) on large-scale data.
Previously: 12 years as Research Science Engineer at the University of Washington (I-LABS) – predictive modeling on 306-channel MEG time series with published effect sizes, co-development of the international MNE-BIDS Python ecosystem (JOSS), and multi-year NIH/NSF-funded pipeline engineering ($2M+). Postdoctoral work at Children’s Hospital of Philadelphia: blinded ROC validation of a language-impairment classifier (AUC 0.86; sensitivity/specificity on N=78) with mixed model inference.
Core tools: Python (pandas, NumPy, SciPy, scikit-learn), R, SQL, Git; Quarto & R Markdown for reproducible reporting.
Open to senior or staff Data Scientist or Machine Learning Engineer roles focused on end-to-end modeling, evaluation, and reliable delivery – not vaporware.