The family health OS — four surfaces, one loop.
A worried, health-building adult child abroad and an aging parent in India, joined by an always-on AI and a family doctor. Below: how AI is woven through, the ecosystem on one canvas, and lo-fi wireframes of every surface. Companion to the market-entry research.
Six roles, one rule.
AI is the connective tissue, not a single "chatbot" screen.
- The watcher (invisible) — reads every reading, writes the daily "Mom is okay" line, surfaces only what matters.
- The explainer (child chat) — "why was her sugar 184?" → context + reassurance + an escalate button.
- The pattern-spotter (proactive) — "her weekend sugar runs high — she skips the Sunday walk" → asks permission to act.
- The voice companion (parent) — warm, spoken; the ambient / Alexa layer for someone who won't tap menus.
- The doctor's copilot — drafts the weekly summary + flags for the doctor to approve (leverage on the scarce resource).
- The self-coach (child "Me" mode) — personal nudges for the caregiver's own health.
The AI screens themselves are sketched within the Parent and Child app sections below (the explainer, the pattern-spotter, the voice companion) and in the Doctor panel (the copilot).
How it all connects.
Device → parent app → AI → child app & doctor → care-manager visit → proof back to the family. Everything routes through the intelligence layer.
Ambient, dignified, never "surveilled."
For the aging parent: big type, three tasks max, an always-present Call + SOS, reassurance baked into every reading, and connection that makes them feel loved, not watched. Includes the AI voice companion.
Anita
feeling today?
today
The caregiver's hero app — Family + Me.
Two modes in the bottom tabs: Family (care — is Mom okay, vitals, alerts, doctor, accountability) and Me (build — the caregiver's own 20-year health-wealth). Includes the AI explainer and pattern-spotter.
adherence
AI drafts, the doctor decides.
A web panel, not a phone app. AI-flagged patients float to the top of the queue; the AI-drafted summary and suggested dose change both sit behind an Approve/Edit gate. This is what lets one doctor safely hold a panel of 24+ families.
The trust layer, on the ground.
The on-ground person in India (Priya, in Kochi). Task-driven: today's visits, an in-home checklist, and proof-of-care that syncs to both the doctor and the family — turning the corridor's "can't verify delivery" problem into verified, delivered care.