A decade of screen-based content — and a new platform with no screens at all.
Business
- Be on Vision Pro at platform launch
- Reuse a decade of existing content — no rebuild
User
- One entry point for three content types
- Zero learning curve in an unfamiliar medium
Technical
- Gaze comfort limits on size and distance
- Legacy formats rendered at real-world scale
No design patterns
visionOS was months old — every interaction designed from first principles.
Three content types
VR, 3D, and AR each behave differently in space — one mental model had to hold them.
Comfort is physical
Design errors cause eye strain — only testable on-device, in the headset.
Solo in a new medium
Every call was mine, with no precedent to lean on.
I owned the product end to end — from strategy to on-device prototyping.
Owned
- Product Strategy
- Spatial UX/UI Design
- Interaction Model
- Prototyping & Validation
Collaborated with
- Product Manager
- Platform Director
- App Engineering
Three calls that defined the product.
Hierarchy by depth, not borders
2D vocabulary — cards, borders, shadows — means nothing in space.
A three-tier glass system: panels at arm's reach, content at viewing distance, environment behind.
Distance is the first thing users perceive — before color or type.
Became the foundation for all spatial UI in the app.




Portals as spatial URLs
A file browser felt like operating a computer inside a headset.
Spatial doorways summoned by code, URL, or title — walk toward one to enter.
Spatial computing should feel like navigating space, not operating software.
First-time users entered content with zero onboarding.
Never jump cut between worlds
Switching between VR, mixed reality, and UI risks disorientation.
Every mode change fades through a translucent glass state.
Users must always answer “where am I?” without thinking.
Zero disorientation reports in on-device testing.


Spatial interface
A complete spatial design language — materials, distances, gestures — built where no patterns existed.
Portal navigation
Discovery as physical movement — doorways instead of menus, presence instead of pages.
Cross-ecosystem rendering
Three legacy content formats living together in one native spatial app.
The trade-off — I prioritized platform-native conventions over brand-specific visual language. On a new platform, meeting user expectations mattered more than differentiation; brand expression is a follow-on layer once the interaction model is validated.
What I'd improve — the experience was designed for a single viewer, but property walkthroughs are inherently collaborative. Shared multi-user sessions are the highest-value next investment.
Next time — I would move to on-device validation earlier. Spatial comfort and scale cannot be judged in flat mockups, and earlier headset testing would have meaningfully shortened the iteration cycle.

