AirAlign

Context

Posture apps often shout. They buzz, they interrupt, they add another notification to an already noisy day.

Challenge

Use motion sensors people already wear — AirPods — to help them notice posture without adding stress.

My role

Concept, interaction design, motion pipeline, prototype engineering.

Process

I sketched how the feedback loop should feel before writing any code, then iterated on thresholds and cadence until the experience felt supportive rather than punitive.

Design exploration

I explored haptic patterns, timing curves, and quiet visual states. The best version was the one users forgot was running until it gently reminded them.

Technical execution

Built with CoreMotion and the AirPods Motion API. Real-time orientation is processed into simple, human categories.

Final solution

A single running state, a subtle nudge when needed, and a calm visual that never asks for attention it hasn't earned.

Outcome

A working prototype that reframes posture feedback as an ambient, respectful interaction.

Learnings

The interface is not always the screen. Sometimes the interface is a small tap you feel in your ear at the right moment.

Key product decisions

• Use AirPods motion data as the primary sensor input.
• Convert sensor data into feedback that is felt, not read.
• Keep alerts calm — no red, no shouting, no streaks to lose.
• Prefer restraint: fewer signals, chosen carefully.