Chroneering
Chroneering AI40 is a cross-platform React Native app paired with the Nord A1 Priority Watch, a digital gatekeeper that filters notifications through AI so users only receive what matters. At Bytemage I delivered 75+ screens across mobile and web, built NestJS backend services, and wired WebSocket and BLE pipelines between smartwatch firmware, mobile apps, and cloud AI/LLM microservices.
The challenge
When we joined, only ~10% of the app was built and most critical screens were missing. BLE watch connectivity was unstable, AI notification filtering was non-functional, backend APIs and cloud integrations were incomplete, and unreliable pairing made real user testing impossible.
Key features
- BLE pairing, calibration, and real-time sync between app and Priority Watch
- AI-powered notification filtering by content, urgency, and user preferences
- Watch pushers and hands synchronized with app functions for alerts and priority ranking
- User onboarding, consent flows, and granular notification controls
- Secure data collection for notifications, app usage, motion, and activity tracking
- Cross-platform React Native app for iOS and Android with OS-specific workarounds
Architecture
Notifications are captured on the phone and forwarded to the watch via BLE. Each notification is split into data packets, decoded, and verified before AI microservices analyze source, content, and severity. Ranked alerts are sent back to the watch only when relevant. The React Native client communicates with cloud APIs for user config, usage stats, and AI-driven filtering, with feedback loops that improve prioritization over time.
My role
Full stack engineer who architected the 75+ screen React Native app and companion web experience, built NestJS backend APIs, and delivered WebSocket and BLE communication with AI/LLM microservices for intelligent notification filtering.
Challenge
Stabilizing BLE watch connectivity on both platforms while iOS sandboxing restricted alarm syncing, and completing 42 missing screens without breaking partially implemented pairing and calibration flows.
Solution
Rebuilt the BLE layer with robust reconnection and state recovery, resolved OS-specific limitations for smooth app-watch interactions, finished critical user flows end to end, and integrated AI microservices so notifications are filtered and ranked before reaching the watch.
Technology stack
Frontend
Backend
Database
APIs & Integrations
Dev Tools