Biography

I am a second-year Computer Science Master’s student at Loyola University Chicago. I come from an interdisciplinary background; I majored in pre-law during my undergraduate studies before pivoting to CS after teaching myself to code sparked a lasting passion and curiosity. I am deeply fascinated by the semantic representation of information in high-dimensional mathematical space, and my ongoing research reflects this interest. I have built a modular architecture centered around detecting drift in data streams relative to a desired baseline. I enjoy applying this framework to a range of challenges within and beyond machine learning, including hallucination detection in large language models, document collation, and instrument calibration in automated data collection pipelines. Looking ahead, I hope to contribute to equitable and trustworthy AI research, pursue a PhD, and develop scalable, reproducible research software that supports open science and responsible innovation.

Academic Status

Masters Student - 2nd

Research Area/Department

Computer Science; Data Science; Machine Learning/AI

Major/Specialty

Computer Science with a focus on machine learning/AI and data science.

Degrees Earned or in Progress

BS in Prelaw/Criminal Justice at Southern Illinois University (graduated May 2023) MS in Computer Science at Loyola University Chicago (in progress)

Academic Preparation

I have completed (or will soon complete) a number of relevant courses, including but not limited to: - Machine Learning - Database Management - Human-Computer Interaction - Intermediate Object-Oriented Programming - Ethics in Computing - Theory of Programming Languages

Research/Publications

I have recently presented research on my software that detects anomalies in automated data collection pipelines at the US-RSE 2025 conference in Philadelphia. The research was published on Zenodo in poster form; a full preprint is in progress and will be published on arXiv soon. Lyons, H. (2025). A Modular Architecture for Detecting Anomalous Data Trends in Research Systems. US Research Software Engineering Conference 2025 (USRSE'25), Philadelphia, PA. Zenodo. https://doi.org/10.5281/zenodo.17267058

Research/Academic Interests

My research interests revolve primarily around the representation of data in high-dimensional mathematical space. Over the past year, I’ve been building a modular architecture that segments this space to establish a ‘ground truth’ of meaning for automated systems of all kinds. I’m passionate about applying this core architectural concept to diverse challenges in machine learning and beyond. I enjoy both developing theories and building modular, elegant software to test and refine them. My big-picture goal is to make automated systems across all domains more trustworthy, efficient, and robust.

Computational and Data Science Areas

Applied Computer Science; Artificial Intelligence and Intelligent Systems; Computer Science; Informatics, Analytics and Information Science; Other Computer and Information Sciences; Training

Motivation

I'm interested in participating in this program and joining this community because I am passionate about making impactful contributions to the national AI infrastructure and the field as a whole. I believe that the tech industry, especially the AI sector, can always be more ethical, more transparent, more efficient, and more environmentally aware. My goal is to push innovation in these directions. I have a strong desire to connect my research in data pipeline supervision to real-world research teams and their automation challenges. I would also like to expand my horizons and make connections in the field to help solve problems that I never would've thought about on my own. I was recently given the opportunity to travel to and present at the US-RSE 2025 research conference through Sustainable Horizon Institute's Building Engagement program. It was an incredible experience that cemented my passion for the research I'm doing and my appreciation for the community I've found through the program. I deeply value the institute's commitment to elevating underrepresented voices in the industry. I share the belief that diversity in perspective leads to better, more robust science, and I would like to continue contributing to that culture. Through this experience, I hope to continue to develop my skills in leadership, communication, and scientific collaboration by taking advantage of the invaluable mentorship and networking opportunities.

Lightning Talk Title

Supervisor Architecture for Semantic & Statistical Drift in AI/ML Systems

Keywords (Maximum 20 words)

hallucination; interpretability; drift; supervision; embeddings; centroids; modularity; architecture; LLMs; robustness; anomaly; detection; pipelines; infrastructure; reproducibility; reasoning; signals; alignment; safety; narratives