Hey, I'm Kennan!
I started programming on calculators with TI-BASIC back in high school, and I've been addicted ever since. Outside of my studies, I'm an avid cyclist, musician, linux enthusiast. Prior to college, I was an active competitor in the Rubik's Cube speedsolving community. During the years when I was actively competing, my highlights included a North American Record, 90 podium finishes, and a top global rank of of 7th and 12th for the 2x2 and 3x3 Rubik's Cubes, respectively.
Currently, I work on graduate machine learning research in the CWRU Department of Computer and Data Sciences under Dr. Soumya Ray. My thesis is focused on accelerated training in communicative learning networks using dynamic structure adaptation. After graduation, I'm transitioning to full time work on Software Engineering and Data Science at the Johns Hopkins University Applied Physics Laboratory, where I work on automated systems for biological threat assessment and full stack software engineering with Angular, Kotlin, and Java Spring.
- Designed and graded course assignments and exams
- Taught supplementary lectures to reinforce course material
- Hosted weekly office hours to answer questions and provide feedback on theoretical written work and programming assignments
- Designed random forest, Bayesian network, and deep learning models for viral and bacterial threat classification using Scikit-Learn, Tensorflow, and Keras
- Implemented a data processing pipeline using dna2vec to perform feature extraction and dimensionality reduction from sequenced DNA
- Contributed to a large scale Angular application to provide an online learing approach to automated document tagging and classification
- Worked in a Kanban development environment to quickly and effectively produce thoroughly documented, tested software for contract sponsors
- Contributed to an Android application written in Java and Kotlin to implement an attachment cache mechanism, reducing form upload time by as much as 75%
- Developed a web application with spring boot backend, Angular frontend, and Selenium unit tests, utilizing an internally designed UI library to deliver Elasticsearch social media analytics
- Developed a job scheduling application from scratch in Angular and designed an algorithm to optimize job scheduling to increase data throughput.
- Designed a Postgres database model to store jobs and their associated data, and built a corresponding REST API to allow application interaction.
- Created an Amazon AWS management server to create and destroy EC2 instances to efficiently allocate funds and expedite job processing.