SpecsModelExplore
Overview
Actors
External Systems
IoT Platform
Edge
Ingestion
Stream Processing
Stream Processor
Kafka ConsumerWindow AggregatorAnomaly Detector
Batch Processor
Device Management
User-Facing Apps
Platform
Stream Processor
Pipeline
Internal structure
Components
API
Related workflows
SpecsModelExplore
Stream Processor
OverviewGuideStream Query APILinks and Communications

Stream Processor

The Stream Processor is the real-time heart of the platform. It reads the telemetry firehose from Kafka, maintains windowed state in Flink, and serves the resulting metrics and anomalies over a small read-only query API.

Why stateful streaming

Aggregates and anomaly scores depend on recent history per device. Keeping that state in Flink (checkpointed, partitioned by device) lets the pipeline answer “what is normal for this device right now?” without round-tripping to a datastore on every event.

Pipeline

Internal structure

Could not build diagram for model "example-iot".

Components

Kafka Consumer

Reads telemetry topics and feeds the pipeline.

Window Aggregator

Computes tumbling and sliding aggregations per device.

Anomaly Detector

Scores aggregates and emits alerts on outliers.

API

The processor exposes a read-only query surface for live metrics, per-device windowed aggregates and detected anomalies. The full contract is embedded on the Overview tab.

Related workflows