Hello! I’m Randy, and I’m in my fourth year studying Computer Science at Georgia Tech with the Intelligence and Devices threads. Previously, I attended the University of North Texas through its Texas Academy of Mathematics and Science.
Currently, I am building a startup called Coordinated Care Inc. with Tee Faircloth to build a webapp-based service to coordinate the transfer of swing bed patients. Often, these post-acute rehab patients transferred from large urban hospitals to smaller hospitals find that they are in facilities better equiped for dealing with day-to-day recovery problems and are closer to their families. Also, medicare pays more to these smaller hospitals for rehab care. We envision that our app will help save small rural hospitals from having to shut down in today’s rural healthcare crisis.
Last summer, I worked for Salesforce’s Mulesoft division on the Cloudhub team, which is the AWS provisioning service for Mulesoft’s main software, Mule. I worked on Cloudhub’s main console, adding API endpoints to edit the console’s internal settings. This way, engineers from both the Cloudhub team and other Mulesoft teams could quickly view settings and change them. Creating these endpoints required extensive searching through code to find possible side effects of changing settings and writing integration tests to verify the correctness of my code.
Last Spring, I was a TA for both the undergraduate and graduate sections of CS 4641/7641 Machine Learning under Dr. Charles Isbell. In office hours and the online Q/A forum Piazza, I spent time answering student’s quetions about fundamental ML concepts, how to use software such as scikit-learn and pandas, and how to make a good/fair ML experiment. While TAing, I recognized that students wanted more help outside of the class’s standard once-a-week office hours, so I ended up organizing a subset of the TAs to hold office hours before test days and paper due dates. The bulk of my hours were spent grading tests and papers on supervised learning, randomized optimization, unsupervized learning, and reinforcement learning.
In Spring and Summer of 2018, I worked at Sandia National Laboratories, working on time series data from hardware. I build a tool that allowed operators to query and filter data, graph it, generate descriptive statistics, and build forecasts with both the regression-based method ARIMA and recurrent neural networks.