Useful expertise is often difficult to articulate and distributed across people working on different problems. I study the interactional work that makes this knowledge available, and design AI that helps collaborators prepare to learn from and support one another.
Understanding coaching expertise
CSCW 2023 · Empirical study. In a university incubator, novices often struggled to recognize whose experience was relevant, ask for useful help, or translate advice from another venture to their own. Through analysis of 24 coaching sessions involving three coaches and 30 novices, supplemented by coach interviews and communication observations, my coauthors and I developed a cognitive model of one-to-many coaching.
Coaches translated project-specific experiences into transferable lessons, connected novices to relevant people, questioned assumptions, and helped articulate requests for assistance. The study identifies the practices that make distributed expertise usable across a learning community.
Evey Huang, Daniel Rees Lewis, Shubhanshi Gaudani, Matthew Easterday, and Elizabeth Gerber. Intelligent Coaching Systems: Understanding One-to-many Coaching for Ill-defined Problem Solving.
AI that helps us help each other
CSCW 2025 · System design and exploratory deployment. Building on this model of coaching expertise, I designed a proactive AI system that helps entrepreneurs surface assumptions and neglected risks before meetings. Mentors receive project context and suggested coaching strategies, and can revise the expert model guiding the AI’s questions.
After iterative prototyping, we deployed the system with one mentor and 11 novices, each preparing for one real coaching meeting. Novices reported identifying overlooked risks and reconsidering their priorities; participants described subsequent conversations as more focused, intentional, and in-depth. Incomplete context sometimes required additional verification. The contribution is situated evidence about preparing people to collaborate, with long-term effects on learning still to be studied.
Evey Huang, Matthew Easterday, and Elizabeth Gerber. AI That Helps Us Help Each Other: A Proactive System for Scaffolding Mentor-Novice Collaboration in Entrepreneurship Coaching. Open-access paper · System code.
Learning to collaborate
CSCW Companion 2023 · Workshop proposal and research agenda. With an interdisciplinary group of collaborators, I developed an agenda for helping people build collaboration skills through complex work. The paper considers how organizational routines, relationships, and technologies support seeking help, articulating needs, and coordinating expertise. It motivates a continuing question across this stream: how does technological scaffolding affect the capabilities people develop over time?
Supporting Workers in Developing Effective Collaboration Skills for Complex Work.
Early work on tacit judgment
CHI 2020 · Computational representations and interactive tools. With Sarah Sterman, Vivian Liu, and Eric Paulos, I studied how people interpret literary style, using crowdsourced comparisons to model aspects of difficult-to-articulate judgments. Interactive tools helped users explore patterns and reflect on their interpretations. This early work established my interest in making tacit judgments available for inspection while preserving a role for human interpretation.
Interacting with Literary Style through Computational Tools.