Phillip Mason

Developed and maintained full-stack applications, contributing across front-end (React/TypeScript), middleware, and back-end (Python, SQL) systems. Adapted and integrated open-source projects into proprietary APIs, containerized services with Docker, and deployed solutions on AWS using DynamoDB and SQL databases. Collaborated on team and individual projects using GitHub version control, ensuring clear communication and efficient workflows in a fast-paced startup environment.

Enhanced billiards gameplay decision-making by developing a deep learning agent using reinforcement techniques, resulting in more effective strategies and intelligent gameplay. Integrated AI with a C++ physics engine in order to train and create a robust learning environment for the AI agent.

Improved SSD drive failure prediction by revitalizing outlier detection with research into LSTM autoencoders as a means to attempt to enhance accuracy and reliability using time series data. Pioneered deep learning research for failure analysis, steering the team towards advanced detection strategies and innovative approaches.

Elevated Codenames AI gameplay by refining capabilities and narrowing the competitive gap with human players, resulting in a more sophisticated AI experience. Automated AI performance comparisons with ChatGPT, streamlining benchmarking and ensuring regular, accurate progress assessment.

Enhanced the capabilities of a social network AI agent by independently teaching myself the necessary methods and techniques. Modified the original agent with a focus on improving its accuracy to reality, demonstrating a strong ability to learn and understand the workings of the initial agent. Implemented new functionality to further improve the performance of the social network AI agent.

Streamlined SSD failure diagnostics by designing software tools, achieving faster resolution times and enhancing customer satisfaction. Developed automated test code to enhance testing processes, reducing errors and analysis time, and boosting overall testing efficiency.

I assisted hundreds of students between Fall and Winter semesters with both understanding and efficiently coding the various projects taught in their computer science classes. This roll was predominately assisting in C++

Conducted in-depth statistical analysis of SSD failures, significantly improving product reliability and reducing customer returns. Enhanced system efficiency by optimizing data interpretation and code validation, leading to faster deployment cycles and improved application performance.