Family In The Loop: Enabling Family Involvement and Person-Centered Dementia Care at Long-Term Care Facilities with Collaborative AI Tools

Person-centered care provides a critical framework for dementia support in long-term care facilities, emphasizing attention to each resident's evolving needs, personal history, preferences, and values. Achieving this level of personalized attention depends on coordination among staff and active family involvement to record and share this essential context. However, fragmented communication, caregiver burnout, limited resources, and regulatory complexity frequently disrupt the flow of information, creating conditions in which person-centered care becomes difficult to sustain in practice. To address these challenges, we introduce Family In The Loop (FITL), an AI-enabled web platform that enhances continuity and collaboration for person-centered dementia care. FITL automatically synthesizes diverse care data-including video, notes, and records-into contextually relevant, emotionally resonant updates tailored for staff and family members. It offers four interlinked views: ''Reels'' provides families with concise, meaningful video updates; ''Essentials'' supplies caregivers with resident-specific summaries for immediate use; ''Explanations'' presents narrative insights into care decisions; and ''Archives'' maintains a searchable, longitudinal record of care activities. Together, these views enable flexible yet coordinated information sharing, respecting the fluid nature of family involvement and supporting continuity across caregiver roles and shifts. Through an iterative co-creation process spanning two years and seven resource-constrained facilities, we designed, developed, and deployed FITL in real-world dementia care contexts. Findings indicate that FITL improved staff access to critical resident information and facilitated sustained family engagement without increasing caregiver burdens. By surfacing timely, personalized, and meaningful information, FITL strengthened trust between families and staff, reduced misunderstandings, and bridged communication gaps. Our work highlights the potential of AI tools to stitch fragmented care interactions into cohesive, person-centered caregiving ecosystems. We conclude the paper by discussing inherent tensions around automation, oversight, and privacy, and we propose design principles for adaptive, collaborative, and person-centered care technologies.