Cleaned and standardized data from a variety of sources such as wearable biometrics devices and surveys of patient-reported outcomes. Data was collected from a patient powered research network of ~16000 individuals who suffer from some form of Irritable Bowel Disease (IBD). Designed and implemented a flexible server API for querying the dataset. Developed PrecisionVissta, a full-stack, doctor-facing visual analytics application to inform and drive patient care decisions. Followed a cyclical development workflow of high-level brainstorming and visual design, technical implementation, and demoing the product and receiving feedback from end users (doctors). Generalized the core visualization algorithm from PrecisionVissta as a technique called "Periphery Plots" (see publications). The Periphery Plots technique is implemented as an open-source, reusable, and extensible React component. JavaScript: D3, React, Antd, Redux Python: Pandas, Numpy, Flask Technologies: Docker, Webpack

Led development of a data source agnostic 3D network graph visualization tool and query engine for large scale datasets. Implemented server-side rendering (remote procedure call interface between JavaScript and Vaadin) to reduce application load time.Researched general purpose clustering algorithms to capture latent features (represented by 3D xyz coordinates) of high dimensional input spaces. JavaScript: Three.js Java: VaadinPython: NetworkX, Sklearn, Numpy, Pandas

Created a MVC UI and REST API for compute cluster configuration and monitoring (SAS Grid Manager).JavaScript: OpenUI5, jQuery