MAY 30, 2025
Vedant Das Swain, PhD
Distinguished Postdoctoral Fellow
Khoury College of Computer Sciences
Northeastern University
About the Presentation: Digital phenotyping holds promise for mental health care by passively monitoring behaviors through smartphones and wearables. However, patients often disengage from these tools because the “seamless” monitoring feels impersonal, untrustworthy, and opaque. What if instead of hiding the seams, we made them visible and configurable?
This talk presents DYMOND, a study exploring “seamful” digital phenotyping that gives users agency over how algorithms interpret their mental health data. Through human-centered computing and participatory design, we examine how 25 participants with depression interact with a configurable passive sensing system over six weeks. Our research operationalizes the seams — the specific interaction points where users want agency in the passive sensing pipeline. We identified seams across the digital phenotyping pipeline; which context informs the model, how behaviors are interpreted, what the model prediction means, and what is beyond the scope of passive phenotyping. These seams reveal the critical junctures where users seek to align algorithmic interpretations with their lived experiences.
Second, we examine interaction design for seam control — how users experience configuring these interaction points and what new affordances they need. We identify needs for collaborative sensemaking, individual adaptation, and scalability challenges. Together, these findings advance human-in-the-loop digital phenotyping by making algorithmic interpretation a collaborative process between patients and their data.
About the Presenter: Vedant Das Swain is an incoming Assistant Professor at NYU and currently a Distinguished Postdoctoral Fellow at Northeastern University. His research focuses on transforming digital applications, behavioral models, and infrastructures to enhance workers’ behavioral health. Dr. Das Swain aims to build digital work technologies to support wellbeing and promote healthier performance in invisible labor, emotional fatigue, and burnout. His human-centered research approach prioritizes applications that workers want to adopt and continually use. He integrates organizational science theories to conceive new data-driven understandings of worker behavior and explores innovative digital interventions to promote healthier work for knowledge work and emotion work.
Vedant has published extensively in top-tier Human-Centered Computing venues such as CHI, CSCW, and UbiComp/IMWUT. His papers have received prestigious awards, including top honors at CHI 2022 and CHI 2024. He has been recognized with the Gaetano Borriello Outstanding Student Award and the GVU Foley Scholar Award. His research is supported by IARPA, NSF, CDC, ORNL, and Semiconductor Research Corporation, and he serves as the lead PI on grants from NIDA/NIH and Microsoft.