As a software engineer at Flatiron Health I work on data pipelines which unify and ingest data from several different practice management systems for use in a data warehouse supporting reporting and analytics. I additionally work on a full stack ASP.NET and React web application, developing features which enable medical billers to organize and manage the billing process. In order to support these various applications, I use Terraform and Ansible to manage cloud infrastructure in AWS.

Worked on developing a wide variety of editing algorithms for light-field photography. The major focus was training a CNN to perform consistent intrinsic image decomposition on light-fields. This included generating architecture in Tensorflow and creating synthetic training data with Blender.

During my internship at Flatiron health I produced a proof of concept migration of the existing monolithic data pipeline for packaging client facing datasets to run on a distributed airflow instance to allow for better reproducibility and scalability.

Programmed analytic solutions for property and behavior prediction in the small molecule space. Projects I was responsible for included, managing AWS based cluster computation of 3d electron structure calculations, building graph convolution based machine learning methods for direct prediction, and programming a quantum approximation library to produce cost effective molecular structure approximations.