Encoding/Decoding Complexity

Spearheaded machine learning, evaluation, and probabilistic forecasting. Developed interactive R Shiny web applications updated daily. Managed fail-safe data ingestion pipelines handling 0.5B-1.5B rows.

Led the Outbreak Management SQL Server Team. Produced complex reproducible reports 7 days a week using R, SAS, and SQL. Developed data wrangling pipelines for all Los Angeles County School COVID-19 data.

Leveraged open-source natural language processing libraries to build tabular time-series from unstructured text. Applied AI for predictive models and contributed to the implementation of parameterized models.

Led a distributed team of data scientists on an NLP project concerning COVID-19. Presented at the 2020 Joint Statistical Meeting (JSM) on supporting emergency response clients.

Mediation, conditional process modeling, growth curves, multinomial regression, and mixed effects.

Dimensional reduction, dealing with missing data, visualization, and time-series data wrangling.

Leveraged supercomputers for Natural Language Processing. Presented on H2O Driverless AI and ggplot2.

Fit generalized low-rank models, Bayesian mixed effects models, and gradient boosting machines for longitudinal responses to breast cancer treatments.

Built cognitive tasks, ran experiments in accordance with human research ethics, and analyzed data via reproducible pipelines for cognitive science.

Served as a resource and community adviser for residents.

Led operational teams to support business objectives.

Co-founded a San Francisco-based health and fitness startup.