Research
My lab studies people in the texture of their everyday lives. We use signals from phones, wearables, and online platforms to understand human behavior and mental health as it actually unfolds. The lab is called SONDER (Sensing, Observing, aNd unDerstanding human ExpeRience), after the word sonder: the realization that every person around you is living a life as rich and complicated as your own.
Research Themes
Understanding human behavior in everyday life. A lot of what we do is descriptive science: using passive signals from phones, wearables, and online platforms to track how people actually live, work, and adapt over time. This includes longitudinal studies of college students across four or five years, how first-generation students adjust to a new environment, how information workers respond to job promotions and burnout, and how all of us shifted during the pandemic.
Representative work: College Experience Study (UbiComp 2024 🏆) · First-Gen Lens (UbiComp 2022) · COVID Student Study (CHI 2022) · Burnout in Cybersecurity Incident Responders (CSCW 2024) · Differentiating Higher and Lower Job Performers in the Workplace (UbiComp 2019)
Detecting mental health from behavioral signals. Building on the descriptive work, we develop AI models that try to detect mental health states (depression, anxiety, suicidal ideation, burnout) from passive behavioral data. We’re equally interested in where these models work well and where they fall short, including questions of cross-dataset generalizability that the field has tended to underinvestigate.
Representative work: MoodCapture (CHI 2024) · GLOBEM (UbiComp 2023 🏆) · Speech-based Suicidal Ideation Detection (UbiComp 2023) · Time2Lang (CHIL 2025) · The Power of Speech in the Wild (UbiComp 2023) · Social Sensing (CHI 2020)
AI-driven, context-aware interventions. The third strand of our work moves from understanding to action: building systems, increasingly powered by large language models, that deliver mental health and wellbeing support in the moments when people actually need it. Recent work spans contextual interventions for individuals with serious mental illness, AI-driven journaling tools for students and productivity agents for information workers. We’re interested in what LLMs can genuinely contribute to mental health (personalization, contextual reasoning, conversational support) and where they shouldn’t be trusted yet.
Representative work: LENS: LLM-Enabled Narrative Synthesis for Mental Health Sensing (ACL 2026) · MindScape (UbiComp 2025) · Context-Aware Mobile Interventions for Serious Mental Illness (IEEE PerCom Magazine 2024) · From User Surveys to Telemetry-Driven AI Agents (CSCW 2025)
Joining the Lab
I’m looking for students who are technically strong (in AI/ML, data science, or mobile/software development) and genuinely curious about human behavior. The students I tend to work best with are people who enjoy moving between disciplines and care about whether their work actually matters to the people it’s meant for. If that sounds like you, you’re encouraged to apply to UVA’s Computer Science PhD program and mention me by name in your application. You’re also welcome to reach out via email with a CV and a brief note about your background and interests. I get many inquiries and may not be able to respond to every message individually, but I appreciate the time you take to write.
Research Collaboration and Funding
UVA Computer Science uses a rotation system in the first year, where students are funded by the university to explore different research groups. It gives both faculty and students a chance to evaluate fit before committing to a long-term advising relationship. I generally don’t fund students immediately who I haven’t worked with during rotations, unless there’s exceptional alignment. Students who continue working with me productively after rotations receive funding through graduation. Current UVA students beyond the first year are encouraged to take my AI for Digital Health seminar and start collaborating on projects to establish fit. Because this funding structure makes changing advisors after the first year difficult, I encourage prospective PhD students to think carefully about research fit when applying and during rotations.
Why UVA?
While domestic students may already be familiar with UVA’s reputation, I understand that institutional prestige and location can be important considerations for international students: founded by Thomas Jefferson in 1819, the University of Virginia is consistently ranked among the top universities in the US (#4 public, #24 overall in US News). Our Computer Science department is nationally recognized, ranking in the top #15 for AI, Human-Computer Interaction, Software Engineering and Mobile Computing. Located in historic Charlottesville in the beautiful Blue Ridge Mountains, we’re just two hours from Washington DC, offering an ideal blend of academic excellence, natural beauty, and access to major metropolitan opportunities.
Publications
For a full and continuously updated list, please see my Google Scholar profile.
I’m not actively recruiting students at the moment, but if your interests strongly align with the work above, you’re welcome to reach out.