Luma / AI & apps
A small space to notice how your days feel.
A SwiftUI wellbeing prototype exploring personal check-ins, conversational journaling, and on-device assistance.
- My role
- iOS / AI development
- When
- 2026
- Built with
- Swift · SwiftUI · Apple Foundation Models
Explore how it works
Try it yourself ↓One check-in becomes part of a pattern.
Choose a mood and energy level, then add the check-in to a sample week.
A sample week
4 check-ins · latest mood: okay · energy: 2/5
A record makes changes visible. This example draws a history; it does not interpret your wellbeing.
An illustrative reflection interaction. It does not run Luma’s on-device model, assess health, or preserve your entries after leaving the page.
The challenge
Reflection is easier to keep up with when it fits into an ordinary day. Luma explores a lightweight way to record a check-in, write through a conversation, and look back at patterns.
How I built it
Make reflection approachable
The prototype brings check-ins, a conversational journal, and habit-and-pattern views into a SwiftUI interface.
Use native on-device models
The companion service uses Apple Foundation Models with structured streamed responses and model-availability checks, plus fallbacks when the model is unavailable.
Explore continuity
Streak and support-circle concepts investigate how a reflective tool might become part of a routine. The project remains a prototype.
What came out of it
A native iOS prototype connecting personal reflection with on-device AI and pattern visualization.
Sources & project context
- Project brief and source, reviewed October 2026