Fleet operations platform, mobile-first
A regional last-mile carrier running 220 vans across two countries on top of a paper-and-WhatsApp dispatch system. Drivers got their stops as photos of printed manifests; dispatch reconciled exceptions by phone. Stop-time variance was eating margin.
We built a React Native driver app with offline-first manifest sync, signature/photo capture, and turn-by-turn integration; plus a Next.js dispatcher console for live route tracking, exception triage, and bulk re-dispatch. Backend is a Node.js services tier on top of Postgres with PostGIS for geofencing.
Within four months, average stop-time variance dropped from ±9 minutes to ±4 minutes per stop. Failed-delivery rate fell 38%. Dispatchers handle 1.7× more vans per shift with the same headcount.
Tech stack
The whole picture.
Top-level technologies. Specific packages and integrations are listed below.
Architecture
End to end.
- Next.js dispatcher console with live map + WebSocket updates
- Bulk re-dispatch via Server Actions
- Offline-first manifest with conflict-free WatermelonDB sync
- Background location with battery-aware throttling
- Photo + signature capture with deferred upload queue
- Postgres + PostGIS for geofence + ETA computation
- WebSocket gateway for live driver telemetry
Libraries
Packages worth naming.
- ExpoRN shell + OTA updates for fleet rollouts
- WatermelonDBoffline-first manifest storage
- TanStack Querybackground sync orchestration
- react-native-mapbox-gldriver navigation surface
- Drizzle ORMtyped Postgres + PostGIS queries
Integrations
External systems.
- Mapboxrouting, ETA, telemetry tiles
- TwilioSMS notifications to recipients
- Auth0driver and dispatcher SSO
- Datadogdevice fleet observability
Building something similar?
Tell me about the product. I'll come back with thoughts and a first shipping date.