I'm a computer vision researcher working at Amazon as an Applied Scientist where I train, test, and deploy neural networks based on state-of-the-art research in new domains. I graduated in 2022 with my PhD in Computer Science from Colorado State University, where I studied representational similarity and bias in CNNs.
As a PhD student I was advised by Dr. Ross Beveridge and contributed to projects in the CSU Computer Vision Research Group, One of my most significant findings is that face recognition models converge to nearly the same outputs after rotation, even when those models don't share training data or CNN architecture. This research has been applied in the DARPA Active Interpretation of Disparate Alternatives (AIDA) program.
Previously, I contributed to the DARPA Communicating with Computers (CwC) program, which focused on DIANA, an intelligent avatar described here.
I also help with Kat McNeely-White's research in various ways, most recently building a simple tool for running MINERVA2 memory simulations.
I worked both alone and on a team to solve business problems using artificial intelligence. This involved coordinating with stakeholders to gather requirements, exploring and gathering large datasets from multiple sources, and training, validating, and deploying deep models to the cloud. I routinely applied state-of-the-art vision transformers and convolutional neural networks to novel domains. Besides technical proficiency, I also provided effective documentation and communication of challenges and results to non-technical stakeholders.
I co-developed, maintained, presented, and published a multi-modal virtual assistant named Diana. I focused on gesture recognition, user perception, and performance optimization (C#, Python 3, Unity engine, TensorFlow, MS Kinect, funded by DARPA CwC).
I collaborated with CU-Boulder and Brandeis Univ. on the creation of a multi-model embedding-based knowledge base for participation in the DARPA AIDA program. My chief contribution was the extension of my MS thesis work to facilitate cross-CNN face identification and correlation (Python 3, TensorFlow, Java).
Beyond grading, proctoring, and occasional lecturing, I taught small workshop-style labs and provided regular one-on-one instruction. The classes I assisted were intermediate C++ programming and advanced DBMS. At the end of the first class, one particular student told me they “couldn’t have done it without me.”
I defined, designed, developed and maintained a web app for managing child participants and associated parents for the purpose of conducting childhood development research (C#, SQL, .NET/MVC, HTML/CSS/JS).
Working on a small, cross-functional, partially remote team, I co-developed adaptive web-based surveys for the purpose of improving patients’ medical outcomes (e.g. pain management, mobility). I was known as a powerhouse and always pushed for using modern design and development tools. While we all wore many hats, I particularly loved back-end optimization and refactoring (C#, SQL, .NET/MVC, HTML/CSS/JS).
I worked on multiple small teams developing console and web applications using .NET/MVC.
I maintained and expanded computational frameworks in support of computational physics research.
In the Material Science and Technology Department, I assisted with the development of scientific instruments, the processing and analysis of experimental materials, and the proposal and presentation of research (Article).
Coursework includes computer vision, machine learning, AI, and distributed systems.
GPA: 3.97
Including most of a Chemical Physics degree.
GPA: 3.71