RLivN¶
A persistent, warm AI companion for elderly and dementia care — always present, never complicated.
Status: Planning — POC not yet scaffolded Org: Big Nerd Idea, LLC — commercial project built on BNI Foundation open source
Overview¶
Elderly and dementia care is fractured across caregivers, family members, and care facilities. RLivN ("Our Live-In") is a planned consumer AI agent that provides a single, persistent, warm presence on a home tablet:
- Client-facing — the person being cared for just speaks and taps. The agent talks, tells stories, answers questions, surfaces reminders, and keeps the client engaged. No menus, no settings, no tech language.
- Caregiver-facing — a human caregiver or authorized family member configures everything through an admin portal: profile, reminders, medications, contacts, and alert rules. Complexity lives here, never on the client device.
- Offline-resilient — a local Ollama model is planned as a fallback when the internet is unavailable, so the home device keeps functioning under connectivity or grid disruption.
- Privacy-conscious — sensitive conversations are intended to stay on-device where possible.
The design target is usability by anyone from age 3 to 103 with zero configuration burden on the client. AI agents are paired with human clients; all setup is done by caregivers — never by the client.
Planning stage
RLivN is in planning and has not been scaffolded. Everything on these pages describes intended design and concept ([HYPOTHESIS]), not built or validated software. Therapeutic-benefit, safety, carbon, and regulatory claims are all unproven — see the Epistemic Honesty directive.
Sections¶
- Summary — this page
- Operations — care model, market context
- Technologies — planned hybrid inference architecture
- Software — planned stack
- Hardware — device requirements, sourcing, sustainability