Felix, an emergency medicine doctor from Zurich, built shii·haa to explore whether breathing feedback could increase self-awareness rather than compete for attention. Most mindfulness apps gamify breathing, but his approach uses live microphone input to reflect your actual breath patterns back to you in real time.

The system processes audio entirely on-device through three layers: signal processing extracts amplitude and spectral features, a breathing state machine tracks phase transitions with adaptive thresholds, and a quality filter rejects noisy windows. This architecture avoids speech analysis and keeps raw audio from leaving the phone, addressing privacy concerns while handling real-world acoustic challenges like background noise and automatic gain control.

Felix acknowledges smartphone-only detection runs behind wearable-based systems in accuracy, but he's running a validation study against clinical ground truth to measure performance. The open-source documentation details the method and research approach without making medical claims.

The project represents a thoughtful attempt to bridge medical expertise with accessible wellness technology, prioritizing honest feedback over gamification. Early results suggest microphone-based detection can support meaningful self-awareness practices.