• Led technical communication during successful application for FDA approval of Machine Learning-based medical device• Converted proprietary Bayesian ML model to Python using PyMC3 for new product feature• Led algorithm testing and documentation process for proprietary technology, including literaturereview, unit test creation, and results presentation for senior stakeholders• Rebuilt critical ETL pipelines (S3/Airflow/Redshift) after complete backend re-platforming and source data schema and architecture changes• Created in-app visualization prototyping tool for fast iteration through user-facing analysis featuresusing Plotly/Dash

• Led team of eleven Data Scientists deploying ML models to AWS-based prediction microservice• Managed construction of CNN- and RNN-based document processing automation pipeline resulting inmore than $1M annual cost savings• Drove regression-based accounting system overhaul for successful compliance with federally mandated revenue recognition standard using Python-based ETL, prediction, and delivery

• Refined model performance and built preprocessing steps for ensemble regression model delivering 98% accurate revenue forecasting• Assembled template RShiny dashboard for daily cross-team product metric analysis• Piloted automated rule generation data mining algorithm for medical charge capture prediction

• Analyzed user actions from transactional data to drive interface overhaul effort• Delivered urgent ad hoc analysis of revenue and client collections trends in support of new marketingefforts and investor presentations

Research coordinator for various cognition-related studies in the UNC HIV/AIDS research program.

Starks Biology Lab