I am an AI research & applications engineer. My primary area of expertise is in machine learning applications & language technology
Engineering
My engineering experience includes core machine learning research & machine learning applications engineering, MLOps, LLM systems engineering, Full-stack web & mobile development. I have worked in industries including policing, healthcare, wildlife conservation, and sports technology.
Teaching
As an adjunct professor at Benedictine University, I taught several undergraduate courses in topics ranging from artificial intelligence, machine learning, and data science.
Research
My research work is primarily in deep learning, NLP, computer vision, and multi agent systems. Most of my research experience was at Northwestern University and in collaboration with the C3 Lab and the IVPL Lab.
Research Topics and Projects
-
Data-Efficient Image Segmentation (2022) - WildWeDeveloped an image segmentation pipeline for a wildlife nonprofit using self-supervised and weakly supervised training, significantly reducing training cost.
-
Multi-Agent Reinforcement Learning (2022) - Northwestern UniversityDeveloped a plan-sharing and collaboration method for multi-agent reinforcement learning.
-
Sentence-embedding Narrative Structure Analysis (2022)Analysis of story plot trajectories in embedding space revealed measurable patterns as plots progressed from setup through climax to conclusion.
-
Video Super Resolution w/ ViT (2021) - Image Video Processing LabDeveloped deep learning architectures for video super resolution (VSR) based on vision transformers (ViT) and CNNs.
-
Narrative Generation Architecture (2021) - Northwestern UniversityCollaborated with Northwestern C3 Lab to develop a narrative generation architecture based on an LLM and narrative planning system.
-
Novelty Generation w/ Neural Cellular Automaton (2021)Explored methods for measuring and generating novelty and open-endedness using cellular automata.
-
Adversarial Defense VAE (2019)Researched variational auto-encoder (VAE) methods for defending image models against adversarial attacks by reconstructing input images to remove perturbations.
-
Style Transfer with VAE and GAN Models (2018)Compared an ensemble of VAE models and a GAN architecture for drawing-to-photo style transfer.