Unicare
Designing an intelligent ecosystem for Canada's healthcare crisis
PROBLEM area
A system failing from every angle
94%
Of nurses reporting burnout symptoms
33 hours
Peak ER wait for a bed in Ontario
54%
Year-over-year surge in wait times
0
National systems for shared lesson-learning
DESIGN GOALS
What I aimed to achieve
Reduce burnout
Support healthcare workers with AI-assisted decision support and team-based care so they can work to their strengths, not just their limits.
Shorten patient waits
Give patients self-triage tools, health insights, and recommendations to help mitigate ER over-reliance.
Empower with data
Give management real-time visibility into team performance, and create a de-identified data brokerage for national lesson-learning.
Unify the experience
Design one cohesive ecosystem so every stakeholder benefits from the same intelligence.
DECISION 1
Three products that share one brain
Instead of a single tool for a single user, I designed an interconnected platform. A patient portal, a bedside decision-support app, and a management dashboard, all pulling from the same data backbone. Improvement in one creates relief across all three.
DECISION 2
Healthcare workers needed a co-pilot, not another clipboard
Literature repeatedly called for team-based care — distributing tasks by skill, not hierarchy. I built the bedside app around a "huddle" workflow: workers collectively assign roles at shift start, then receive AI-powered guidance throughout the day. Every recommendation shows its reasoning, so clinical autonomy is preserved.
Voice assist was critical — healthcare workers' hands are rarely free. The app also surfaces real-time IoT patient alerts, catching anomalies before they escalate.
DECISION 3
Give patients contextualized health plans
Research showed patients treating ERs as a "one-stop shop" — not by choice, but because nothing else gave them clarity. I designed the patient portal as a health home: records, connected devices, appointment insights, and most critically, a self-triage flow that walks users through symptoms, asks targeted questions, then recommends a course of action — clinic, telehealth, or ER.
The goal isn't replacing medical advice. It's giving people the understanding that stops unnecessary ER visits from happening in the first place.
DECISION 4
Hospital data was trapped. I turned it into shared intelligence.
The most striking research finding: when a hospital solves a problem, that knowledge stays local. No national system exists for lesson-learning. The admin dashboard surfaces real-time efficiency metrics, but also enables de-identified data brokerage, hospitals anonymize operational data and share it, or sell it to research organizations.
RETROSPECTIVE
Results & insights
View full PDF
ACIDO Rocket Finalist
Nominated by my thesis advisor Alexander Manu to compete in the Association of Chartered Industrial Designers of Ontario's Rocket program. Selected as a finalist.
3 products, 5 stakeholder groups, 32 sources
An interconnected ecosystem prototyped, card-sorted, and validated with SMEs. Every feature traces back to empathy maps, value proposition canvases, and a 32-article literature review.
What I'd measure next
Reduction in ER misuse rates, healthcare worker burnout scores (Bradford Factor), patient satisfaction with health comprehension, and revenue from de-identified data brokerage.
Next steps
Feasibility testing, design system, coded MVP, hospital partnership for on-site iteration.
