AUXO: IoT Fleet Platform at Telecom Scale
Real-time telematics ingestion at 5,000+ vehicle scale: event-driven processing over MQTT and RabbitMQ, and a legacy-to-cloud migration on AWS.
01
Context
AUXO is Telenor Pakistan's end-to-end IoT fleet and asset management platform: real-time vehicle tracking, geofencing, and fleet analytics for 5,000+ connected vehicles. The product family extends beyond fleet into Cold Chain monitoring, Genset monitoring, and Asset Management, alongside an M2M SIM Management Platform.
Telemetry arrives continuously from devices in the field over MQTT and raw TCP — a firehose of position, status, and sensor events that has to be ingested, processed, and surfaced to customers in real time.
02
My Role
I worked on AUXO as a Full Stack Developer in Telenor Pakistan's IoT & M2M team from March 2024 to January 2026, across device pipelines, backend services, and customer-facing features.
- Built features across the microservices + event-driven architecture, with RabbitMQ as the message backbone and device pipelines over MQTT/TCP
- Migrated legacy systems to cloud-native AWS with Docker
- Improved database response times by 30% through indexing and query optimization
- Extended the platform across AUXO Fleet, Cold Chain, Genset, and Asset Management; contributed to the M2M SIM Management Platform
03
Architecture
Devices publish telemetry over MQTT (and raw TCP for some device classes) into ingestion services that normalize events and hand them to RabbitMQ. From there, independent consumers handle tracking state, geofence evaluation, alerting, and analytics — the classic event-driven split that lets a spike in raw telemetry throughput scale ingestion without touching the read side.
The platform runs on AWS in Docker containers, the result of a migration from legacy self-managed systems to cloud-native infrastructure.
04
Key Challenges
- Scale and burstiness: 5,000+ vehicles emit telemetry continuously, and events arrive late, duplicated, or out of order — geofencing and trip logic must tolerate all three
- Legacy-to-cloud migration: moving live device pipelines to AWS/Docker without dropping data from vehicles in the field
- Database performance under telemetry write load — addressed with indexing and query optimization that cut response times by 30%
- One platform, many verticals: keeping Fleet, Cold Chain, Genset, and Asset Management on a shared core without forking the codebase
05
Outcomes
- 5,000+ connected vehicles with real-time tracking, geofencing, and fleet analytics in production
- 30% faster database response times after indexing and query optimization
- Platform supported 500% customer growth over its 2019–2023 period
- Legacy systems retired in favor of cloud-native AWS infrastructure with Docker