I am beginning prerequisite coursework towards applying to DPT programs. I am actively looking for opportunities to shadow physical therapists. I am excited to learn about all specialties, but especially neurology, orthopedics, and pediatrics. I also do some work at the intersection of physical therapy and data science on the side. I am curious to see how data-driven technologies will change the profession, and would like to do my part to make sure those changes are humane and human-centered.

Building data resources and pipelines to support various web applications.

Compared the effectiveness of leak detection and notification policies against behavior change messaging in the East Bay Municipal Utilities District.Modeled municipal water pipe breaks using a self exciting point process framework to support and test break prediction methods in the East Bay Municipal Utilities District.Quantified evaporative cooler water consumption to assist the California Department of Water Resources in making budget variance calculations.Built and maintained SQL databases of hourly energy data spanning multiple years and tens of thousands of accounts.Featurized client activity streams from online portal usage to aid in modeling traction of behavior-change campaigns.

Built and maintained production-level data pipelines to transform terabytes of field-level agronomic data into weekly reports in user accounts.Developed statistical algorithms and heuristics to search gigabytes of field-level data for quasi-experimental spatial features.Processed and featurized large, highly-autocorrelated spatiotemporal datasets.Performed analyses of causal inference on a variety of agricultural practices and products. Built prediction models of economically important temporal variables (such as harvest date) using historical spatiotemporal agronomic data.

While obtaining my Masters Degree at Berkeley I was Graduate student instructor for Stat 20: Introduction to Probability and Statistics. Held discussion sections and office hours. Proctored and graded exams.

Performed meta-analyses using Monte Carlo Markov Chain based bayesian linear models to compare next-generation pesticides with baseline chemicals.Implemented machine learning methods, including K-Nearest Neighbors and Convolutional Neural Networks (using Theano), for classification/regression problems and statistical simulations.Queried and cleansed data from the PUR, CIMIS, and NCEI databases using R, Python, and PostgreSQL.Wrote R scripts to calculate and compare a variety of water-energy metrics of drought-regulation compliance.Implemented predictive algorithms, including a negative-binomial generalized linear model, support vector machines, and generalized additive models, to estimate almond harvest loss for the California Almond Board.Cleansed and performed quality-control of data using R, plyr, dplyr and parallel processing.Led statistical simulations of novel water-meter rates to estimate Utility and customer savings.Performed meta-analyses using Monte Carlo Markov Chain based bayesian linear models to compare next-generation pesticides with baseline chemicals.Implemented machine learning methods, including K-Nearest Neighbors and Convolutional Neural Networks (using Theano), for classification/regression problems and statistical simulations.Queried and cleansed data from the PUR, CIMIS, and NCEI databases using R, Python, and PostgreSQL.Implemented predictive algorithms, including a negative-binomial generalized linear model, support vector machines, and generalized additive models, to estimate almond harvest loss for the California Almond Board.Cleansed and performed quality-control of data using R, plyr, dplyr and parallel processing.