- Researched, designed, built, and deployed machine learning models in collaboration with business partners using Python, Scikit-Learn, and Spark ML- Built hospital readmissions risk model, scoped to create $20M in value annually as part of an intervention program- Created prioritization and scheduling logic for telephonic outreaches to customers, estimated to increase the success rate of the campaign by up to 9% per month- Used constraint programming to create keyword list for call center notes text classification, reducing false positive rate by 40% over existing list while maintaining same true positive rate

- Built and flew an unmanned aircraft system to observe plant life cycles with digital photography, as shown in background photo- Used image processing, curve fitting, Monte Carlo uncertainty analysis, and variance partitioning to analyze data with Matlab, R, and Python- Discovered effects of environment and species on plant activity at unprecedented spatial and temporal resolutions- Presented to specialists and non-specialists at 5 local, national, and international meetings- Led team of 2 software engineers on data science project, using logistic regression in scikit-learn to create real-time predictor of leaf life cycles

- Assisted in teaching classes of 120-350 undergraduate science majors and non-majors in Data Science (3 semesters), Advanced Scientific Computing, and Science and Cooking- Created final project using Python, Pandas and regularized regression in scikit-learn to predict effects of diet on disease; increased model testing score 40% over ordinary least squares- Mentored 2 undergraduate researchers through the Harvard Forest Summer Research Program in Ecology

- Consulted for the New York City Department of Environmental Protection on a real-time warning system for combined sewer overflow and waterway bacteria events- Developed a time series/logistic regression model in scikit-learn to increase accuracy 25% above baseline models currently in use- Coordinated implementation of new model on Department web site

- Developed process-based model of impacts of selective logging on forest carbon sequestration in Matlab- Presented results to organization leaders

- Assisted with data management and analyzed digital photography of grassland ecosystem function- Presented poster at national meeting

- Developed mathematical models of contaminant transport in water distribution networks- Designed and conducted lab experiments for model validation, implemented models for predictive simulation- Received best paper award for oral presentation at the Water Distribution Systems Analysis Symposium

Civil and Environmental Engineering - Hydrology

Algebra II, Calculus, Assistant Band Director

Math