Raztom Lab
Case Study

Zerashi

Zerashi is a quiet companion for Garmin watches, shaped by a small game idea, feedback from friends, the stress of sirens in Israel, and real testing on a physical watch.

BetaA quiet companion for Garmin watches
Role
Wearable product direction, AI-assisted build
Timeline
Prototype with device testing
Stack
Garmin Connect IQ, watch data, companion logic
Status
Beta

The Problem

Many smartwatch apps compete for attention with alerts, streaks, dashboards, and constant measurement. I wanted to explore a gentler idea: a small presence on the wrist that feels more like a companion than another thing demanding focus.

The Constraints

The project started as a small desktop character, then friends suggested it could be more meaningful on a smartwatch. Building for Garmin was new to me, and the work had to move from planning, to simulator testing, to real checks on an older Garmin watch my sister gave me.

The Solution & Build

I used AI to learn the Garmin environment, test different approaches, and build a companion that changes with time, steps, weather, visits, and Calm Mode. The calm feature was shaped by the pressure of war and sirens in Israel: a small character that can accompany the user with a softer tone in stressful moments.

The Impact

Zerashi taught me how different a watch product is from a website: the wrist is personal, interruptions matter, and small decisions affect comfort. The Quick Launch split was a practical product decision: avoid a crash, keep the experience stable, and continue improving from real feedback.