Ontario Health · Health811 · 2022 to present
Designing the conditions for trust in AI supported health navigation
The AI Governance Table adopted my trust framing as its evaluation criteria, and the repositioning of Health811 toward a health concern problem solver is adopted and in build.
Context and stakes
Health811 is Ontario's front door for health navigation, a public service used by millions of residents. When AI symptom assessment entered the roadmap, the risk was treating it as a standalone intake tool bolted onto a service listing.
My role
I mentor senior and intermediate designers across product areas, and I started a cross-program UX and CX community of practice, where leads work through projects in flight and open questions, and where I have demoed AI-assisted prototyping, from a visual to working multi-page flows in Claude Code.
I lead UX for Health811, public facing digital health services and logged in provider and clinician experiences used by millions of Ontarians, and I serve as the invited UX representative on the AI Governance Table, which evaluates where and how AI belongs within a province wide health navigation service.
I partnered with programs and vendor teams to embed WCAG standards into shared design system patterns, holding consistency of application across teams as the quality gate.
The question is not whether the tool can work. It is what conditions have to be in place for it to be trusted, understandable, and safe.
What I built or changed
I reframed evaluation away from model performance and intake efficiency and toward the surrounding experience: plain language safety messaging, clear escalation paths, human handoff points, and accessibility checks. That shifted the broader product conversation too, repositioning Health811 from a service listing model toward a health concern problem solver, where AI is one part of a safe path from describing a concern to reaching the right service.
Concretely, that reframe changed the unit of evaluation. Rather than asking only whether the model returned the right answer, the table began asking what happens when it does not: whether the safety message is in plain language, whether the escalation path is obvious, and whether a person is reachable at the moment it matters. The accessibility checks sat in the same frame, so trust was assessed as a property of the whole experience, not the model alone.
Outcomes and open threads
Two things have landed. The AI Governance Table adopted this trust framing as its evaluation criteria, and the repositioning of Health811 toward a health concern problem solver is adopted and in build. Neither is a shipped AI feature yet, and I am not claiming one.