A high-throughput CQRS system that ingests device telemetry via MQTT, processes it through Kafka, and delivers real-time fleet visibility — all in milliseconds.
Your telemetry data flows through a battle-tested pipeline in real-time.
Mobile devices publish GPS telemetry as Protobuf over MQTT/WSS to the Mosquitto broker.
Vector transforms and routes MQTT messages to Kafka topics for reliable, ordered delivery.
The Go ingestion service writes to TimescaleDB for persistence and publishes to Redis for real-time updates.
The Fleet Service streams positions via WebSocket to the React dashboard for live fleet tracking.
The Go-based ingestion service processes 10K+ messages per second with idempotent writes, dead-letter queuing, and automatic retry logic. Built with hexagonal architecture for testability.
Track every device on a live map with sub-second updates via Redis pub/sub and WebSocket streaming. The Python Fleet Service delivers REST APIs and real-time position feeds.
End-to-end encrypted telemetry with WSS transport, Protobuf wire format, and OIDC-based authentication. Every byte is validated, every connection is authorized.
High-performance ingestion with goroutines and channels
Time-series optimized PostgreSQL for telemetry storage
Lightweight pub/sub for IoT device communication
Durable message streaming between bridge and ingestion
In-memory pub/sub and geo-queries for real-time positioning
Modern dashboard with live WebSocket maps and Tailwind UI
Open the dashboard and see your devices in real-time.
Open DashboardSelf-hosted · Open source · No vendor lock-in