Year: 2025
Type: article
Abstract: Loneliness and social isolation among older adults have emerged as urgent public health challenges, with significant physical, emotional, and economic consequences. In response, this study explores whether emotionally intelligent AI agents can support older adults through sustained, meaningful interactions. We present Seva, a voice-based, GenAI-powered conversational companion designed to reduce loneliness by simulating empathetic, personalized dialogue. The study follows the Elaborated Action Design Research (EADR) methodology, progressing through diagnosis, design, implementation, and evolution phases. This informed the development of a mobile-based app integrating large language models, voice recognition, and personalization modules. Early findings suggest that features such as empathetic tone and memory-aware responses were positively received. This research contributes to the ongoing discourse on AI-enabled companionship and provides early-stage insights into the development of emotionally attuned conversational agents for aging populations.
Authors: Ray, Arindam, Munshi, Archisman
Institutions: University of North Carolina at Greensboro, Thrive
Language: English
Subfield: Demography
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7018612915
FWCI:
Cites:
Cited by:
Related to: