Catch Risk Early, Before It Escalates

CaritaHub's Risk Intelligence system continuously monitors every senior across clinical, functional, and psychosocial domains — surfacing risks, suggesting actions, and keeping your team informed. Human judgement stays at the centre of every decision.

How It Works

1

Collect

Data is drawn from member profiles, IoT wearables, clinical assessments, and psychosocial observations.

2

Score

Risk is normalised and weighted across three domains, with a velocity modifier that adjusts scores based on how quickly a senior's condition is changing.

3

Analyse

Nine AI agents validate the data, detect patterns, and check for clinical appropriateness — with built-in safeguards at every step.

4

Act

Prioritised care plans, escalation pathways, and audit-ready reports are generated for the care team to review and act on.

Three Domains of Risk

Clinical (40%)

CFS as the core baseline. Vital signs including blood pressure, heart rate, and SpO2. Wearable data covering HRV, sleep, and stress.

Functional (35%)

SPPB for mobility and fall risk. Daily step counts and ADL completion. Sleep quality over time.

Psychosocial (25%)

GDS-15 depression screening. MMSE cognitive assessment. Social isolation index.

Nine AI Agents

Tier 1 — Data Collection

Risk Analyst

Scores nine indicators and produces a composite risk assessment.

Health Monitor

Watches clinical thresholds and triggers a re-score when data suggests decline.

Activity Tracker

Detects drops in engagement and flags social withdrawal.

Tier 2 — Clinical Review

Quality Evaluator

Audits every output. Blocks results that don't meet clinical standards before they reach the care team.

Tier 3 — Communication

Notification Coordinator

Routes alerts based on severity: calls for Critical, WhatsApp for High.

Conversation Guide

Adapts tone and communication style to each senior's personality profile.

Post-Analysis

Report Generator

Produces tailored reports for four audiences: management, clinical teams, families, and regulatory bodies.

Care Planner

Proposes interventions based on risk findings. All plans require human sign-off before any action is taken.

Clinical Guardrails

Human sign-off always required
No care plan is changed, and no escalation is triggered, without a care team member reviewing and approving it first.
Override rules
Certain conditions automatically elevate risk classification — CFS ≥ 6, active hospitalisation, or a fall in the past seven days always trigger a High or Critical flag, regardless of other scores.
Continuous improvement
A quality evaluation loop runs on every output to improve accuracy over time.

Talk to Us About Risk Intelligence