Researcher · Builder · Writer
I study what happens to human judgment around AI. I build the systems that keep humans in charge.
- Research that changes decisions. 8+ years leading mixed-methods UX research at IBM, Stanley Black & Decker, and State Farm. This year's studies, a three-phase MaxDiff with ~177 participants and a within-subjects tree test triangulated against ~10,900 real users, are steering product positioning and navigation.
- I build what I study. Five production-shaped apps and a custom AI research-skills suite that cut transcript coding from ~3 hours to 5–15 minutes per session. The automation makes evidence traceable. The judgment stays human.
- Selected accomplishments. Co-author, Design Principles for Generative AI Applications (CHI 2024). Founded IBM's company-wide UX competitive benchmarking program. Writing in Smashing Magazine and UX Booth, with 200K+ readers at renaissancerachel.com.
Selected evidence
A navigation tree test where every participant tried both information architectures: 80% preference for the simplified nav (p = 0.01), triangulated against behavioral data from ~10,888 real users. The simplified navigation was adopted. The method, written up for LogRocket.
Founded IBM's company-wide UX competitive benchmarking program: the framework, the first three benchmarks, and the training for other researchers to run it. The watsonx.data vs. Databricks benchmark won the Q3 Entrepreneur Award and an Outstanding Technical Achievement Award. The approach, on the Awkward Silences podcast.
A custom suite of AI research skills that cut transcript coding and repository processing from ~3 hours to 5–15 minutes per session, with first-pass analysis kept human-led. Packaged and open-sourced on GitHub.
The relationships that make research possible at all: 7 steps to build customer relationships for user research, for Smashing Magazine.