Janu XR
An AI-assisted prototype for helping first-time XR users through fit, eye calibration, and pinch-gesture setup.
A first headset session can become frustrating when fit, calibration, or unfamiliar gestures fail without explaining what to change.
Inside the experience
What someone can actually do
- Move through an eight-step headset setup flow
- Trigger targeted help after three simulated failure signals
- Receive step-specific guidance instead of generic help
- Review sessions through Flow, Insights, and Replay
- Explore simulated gaze-linked guidance in Unity
How the idea took shape
Poor onboarding can end a session before someone reaches the experience the headset was meant to provide.
Janu XR began by examining first-time setup, physical discomfort, unclear gestures, fear of making mistakes, and the gap between what a headset can do and what a new user feels ready to do.
The resulting experience guides setup through eight steps. Rather than interrupting every hesitation, three accumulated simulated event signals trigger focused help for the step that is failing.
A FastAPI service records the experience, Android views make the flow observable, and a Unity prototype demonstrates contextual gaze guidance. The prototype uses simulated gaze and calibration events, not live headset telemetry.
“First-time XR users do not need more generic instructions. They need contextual reassurance when a physical setup step starts to fail.”
Explore how it was built +
I framed the onboarding problem, designed the intervention logic, built the FastAPI service and Android evidence views, and created the Unity gaze-guidance prototype.
- Python
- FastAPI
- Gemini
- Jetpack Compose
- Unity
- C#
Where it could go next