Battling online test cheating by utilizing ML techniques.

Built a custom ETL on AWS to help Comcast understand their data and make their internal events visible to people working in call centers. The goal was to simplify access to the information, make it more understandable, and improve the understanding of what data was available.We created a self service model where other teams could onboard their data to us so it could be categorized, normalized, and shared across the company using our API and UI.We received roughly 200 million events per day through our ingestion pipeline, and 700k requests per day on the read side.

We built a UI capable of visualizing second to second data, or data covering an entire month. We allowed the user to zoom in and out of the data, viewing which different smoothing algorithms. Often we had to visualize 100,000 data points at once and allow the users to interact with them to manage their own power usage.We also build a middleware service which acted as a bridge between the analytical scripts written Python and the UI. It was a Node service that wrapped the Python scripts allowing the statisticians to expose new models without any code beyond their models.

Developed medical focused based web applications. In the early years of the SPA craze, we built an app that allowed doctors to mark on a picture of an eye the damage they saw. This did give me an appreciation for the power of raw data and the benefits it can bring.

Built eCommerce sites for numerous clients.