About

I'm Mike, a Postdoctoral Research Fellow at the University of Michigan School of Information, working with Kevyn Collins-Thompson. My favorite problems sit at the overlap of statistics, natural language processing, and education research: places where no single field has a clean answer. I build methods for analyzing human conversations at scale (data infrastructure, temporal and discourse features) and evaluation frameworks for the generative AI systems now embedded in those conversations.

I came to this work through an interdisciplinary path: a BS and MS in mathematics at Cal Poly, Peace Corps service in Botswana, and a PhD in math education at Michigan with Deborah Ball. Teaching has been a through-line: math and stats departments, advanced math camps, middle- and high-school enrichment. The unusual route is the point. My real expertise is interdisciplinary, and most of what I find interesting lives in the crevices between fields that don't usually talk to each other.

Day-to-day, what actually drives me is writing code to test an idea. I like staying close enough to the data to still tell what the numbers mean, and I like the moments when a method built for one field turns out to work cleanly in another. Most of what I do now uses statistics and machine learning, but I think of myself less as a statistician or ML researcher and more as someone who will pick up whatever tool the problem actually needs.

I live in San Luis Obispo with my wife and our goldendoodle. Outside of research I play guitar and ukulele, hike when I can, get into the occasional racket sport, and meet up with friends for board games and Super Smash Bros. Melee.

On the job market

My postdoc runs through August 2026, and I'm looking for what comes next, open to both industry and academic roles. The common thread I'm optimizing for: building evaluation methods for language models and the human-interaction data they learn from and act on, somewhere I can stay close to the data and the questions. Concretely, that means research-scientist or applied-AI roles (LLM evaluation, conversational and educational data, measurement) or tenure-track / teaching-research faculty in statistics, computer science, data science, or related departments.

If that overlaps with what your team or department is hiring for, I'd love to talk. The fastest way to reach me is email, and my CV is here.

Get in touch

I also take on a handful of consulting and research-collaboration projects each year. Best entry point is email: mikeion@umich.edu. You can also find me on Google Scholar or GitHub.