At present I am between positions and taking some time to perform basic AI research explaining the function of the transformer stacks underlying technologies such as ChatGPT.Previously, at DISCO, I introduced the first commercial deep learning application in the $10 billion e-Discovery market; this product was a major revenue driver for DISCO, which was not previously an AI company. The AI products eventually became one of the core planks of DISCO's IPO prospectus when it debuted on the NYSE in 2021 with ~$100 million revenue. As an experienced technologist and researcher in artificial intelligence, natural language processing, and cognitive architectures with primary focus on deep neural networks for large language models,I have lead machine learning & artificial intelligence efforts at DISCO for 8 years. Most recently, I have worked on a natural language search tool based on large language models. Historically, DISCO AI offers a predictive coding tool for e-discovery that performs iterated active learning on corpora involved in legal proceedings. This tool trains a multimodal neural network to integrate text, image, and metadata in order to predict the likelihood of relevance for user-defined criteria in user-uploaded corpora of 1GB - 10 TB data each.I hold a Ph.D. in Computer Science from the University of Texas, Austin, which is highly ranked in the field of Artificial Intelligence. While there, I worked on optimization theory, especially with respect to evolutionary computation for training neural networks. My postdoc on Robotics & Deep Learning (2013-2015) was funded by an NSF grant with Jürgen Schmidhüber at IDSIA in Switzerland. I am listed as inventor on three patents and two other pending patents and have over a dozen publications in topics including game theory, theory of optimization, functional analysis, neuroevolution, graphical models, and robotics. A monograph arising from my thesis on the analysis of optimization methods was published with Springer in 2020.