ACUITY · AI Health Monitoring · 24hr Challenge · 2026
Determining When Health Signals Require Action
Overview
Modern wearables detect signals — but don't translate them into action
Patients experience symptoms and see elevated metrics, but don't know what they mean or what to do. ACUITY bridges that gap by detecting sustained deviation from a user's baseline and escalating guidance with proportionate urgency: nudge → schedule → urgent care.
The problem
Wearables provide raw signals (heart rate, blood oxygen, ECG) but leave interpretation to the user. Symptoms + metrics still leave patients uncertain: is this serious, or just stress?
The opportunity
Systems can detect sustained baseline drift, classify risk tiers, and guide users toward the right care pathway without alarming or over-triaging.
The constraint
High-risk patients are least likely to self-diagnose. Deployment through insurers and health systems reaches them with integrated pathways, not just app notifications.
Solution
An intervention engine that translates biometric drift into clear medical next actions
ACUITY detects sustained deviation from a patient's baseline and routes them to the appropriate care tier. The system's "when do we speak?" logic is designed as a first-class UX surface.
Personalized baseline
The system learns each patient's normal patterns from initial weeks of monitoring, creating a personalized reference point instead of using population averages.
Deviation detection
Continuous monitoring flags sustained abnormal patterns — not one-off spikes. Magnitude, duration, and symptom context combine to determine escalation tier.
Tiered interventions
Tier 1 (monitoring) → Tier 2 (schedule evaluation) → Tier 3 (urgent escalation). Each tier has proportionate language, confidence signals, and clear next steps.
Core Flows
Four touchpoints that close the gap between signal and action
Discovery, alert logic, scheduling, and provider handoff — the minimum proof of an end-to-end care pathway.
Establish personalized baseline
The bracelet captures baseline vitals across different states (rest, activity, sleep). The system learns normal patterns for this specific patient, creating a personalized reference point for detecting meaningful deviation.
Tiered alert that explains the "why"
Sustained drift triggers a proportionate alert. The card explains risk tier, plain-language interpretation, and confidence signals. Primary CTA is "Schedule evaluation" for Tier 2, with escalation pathways for Tier 3.
Low-friction handoff to care
The product is only as good as follow-through. Scheduling reduces steps, pre-fills context from the alert, and confirms what happens next. Integration with provider systems ensures seamless transition from app to clinic.
Clinician-ready summary
A concise report translates biometric drift and symptom context into medically relevant information. Providers can triage quickly without patients needing to interpret charts or explain their own metrics.
System
Closing the data-action gap requires designing the intervention system
ACUITY's intelligence is not "AI magic." It's a legible system that weighs drift magnitude, duration, and symptom context to route patients to appropriate care tiers with high confidence and low false-alarm rates.
Deviation detector
Continuously compares current signals to personalized baseline. Flags sustained abnormal patterns, not one-off spikes. Accounts for context: activity level, time of day, recent stress.
Logic gate
Routes users to appropriate tier by combining confidence thresholds, risk priors, and user-reported symptoms. Designed for false positives: only escalates after sustained drift, uses cautious language.
Care pathway
Tier 1 (monitor) → Tier 2 (schedule) → Tier 3 (urgent). Each tier connects to real care infrastructure through health systems and insurers, not just app notifications.
Design Decisions
The decisions that made ACUITY credible under time pressure
Each decision came from identifying the core system behaviors that create trust.
Personalized baseline over population norms
Population-average thresholds miss meaningful drift for individual patients. We chose to learn each patient's unique baseline, making the system more sensitive to personal change.
Tiered interventions over binary escalation
A single "alert or not" decision causes either over-triage or missed signals. Three tiers allow proportionate responses: subtle nudge, scheduled evaluation, or urgent care routing.
Insurer deployment over direct-to-consumer
Patients who most need preventive monitoring are least likely to self-purchase wearables. Reaching them through health systems and insurers ensures integration with actual care pathways, not just notifications.
Clinician summary over raw data dashboards
Providers need context, not charts. A concise summary translating biometric drift into medical relevance enables faster triage and higher follow-through rates.
Impact
Reducing the gap between symptom and seeking care
User outcomes
Shorter time-to-action. Lower uncertainty. Alerts are meaningful because they're grounded in personalized baseline, not population thresholds. Escalation only happens after sustained drift, reducing alarm fatigue.
System value
Fewer avoidable emergency admissions. Aligned incentives between insurers, providers, and patients. Scalable care capacity by routing elevated cases to clinics first, not ERs.
Clinical impact
Early intervention on high-risk patients before acute events occur. Modeled outcomes: 30-40% reduction in time-to-care, 1-2% reduction in emergency admissions, $8-12M annual cost reduction for health systems.
Success is measured by whether high-risk patients seek evaluation sooner and whether escalations reach appropriate care tiers consistently.
Reflection
What the 24-hour sprint revealed
Time pressure forced systems thinking
24 hours meant designing proof of a thesis, not a gallery of screens. The minimum viable proof was a closed loop — every design decision followed from the core hypothesis: can we detect drift and route patients to appropriate care?
Infrastructure beats novelty
The wearable is not the story. The story is the intervention engine — the system that decides when to act, how confidently, and where to route care. That's what distinguishes ACUITY from just another health app.