Implemented media security features, including manifest manipulation and segment encryption, for live-streaming media using C++.
Implemented Brotli compression, weak ETag support, and support for “partial combined objects” on Akamai’s
web products. One customer experienced over 30% fewer origin hits, and most experienced up to 10% quicker average delivery times to mobile users after opting in.
Created libraries to build animated, interactive data visualizations using Python, React, D3.js, and WebGL.
Implemented microservices on a Kubernetes cluster that continuously run data science queries from multiple databases and stream the results into an in-memory cache for instant access to new information.
Created a containerized Web application, written using Python and rendered by a React/Redux front-end, to display these data streams inside interactive dashboards and on-demand, procedurally generated reports.
Designed and led development of a scalable data pipeline written in Python that loads historical and real-time web analytics data into a central database while minimizing compute cost and tolerating extraction faults such as rate limiting.
Patched firmware for Tegra's power management processor to remove MISRA C violations, in order to comply with software safety regulations per ISO 26262.
Substantially refactored the codebase to eliminate unsafe debugging routines from the production build.
This internship was completed as part of a Major Qualifying Project (capstone) at WPI during the spring semester of 2019.
