Akamai Resource Timing Data

For the UCSC computer science software development series, I had the opportunity to work on a research project sponsored by Akamai Technologies to analyze resource timing data collected from their network to predict performance improvements gained by applying the available optimization techniques on Akamai’s Content Distribution Network. I wrote Puppet manifests to automatically configure uniform development enviroments, including support for Apache, BIND9, Hadoop, Impala, Pandas, Numpy, and Akamai’s Perl resource timing data analysis framework, as well as contribute to development of Python data analysis scripts for predicting successful performance optimizations.

Community for Me

Community for Me is a web application for finding resources relevant to the communities you identify with. I wrote this web application in Go with a couple of friends during the UCSC Hackathon in 2015, and got to use a collection of technologies including Go, Vagrant, Heroku, Mandrill, and CloudFlare.

Smart Energy Analytic Disaggregation System

The SEAD microgrid project aims to provide analysis of home power grid usage to provide analysis to optimize home energy usage. I worked on the backend data collection server software written in Go, as well as a simple Python 3 API to provide access to the data stored in a Postgres backend, which would be in turn served by the front end web app. Server provisioning was accomplished via Puppet and a couple shell scripts, and setup was tested with Vagrant.