real-time-analytics

Real Time Analytics

A ride-sharing platform needs to compute real-time metrics from a stream of trip events: ingest the raw event, enrich it with driver and rider metadata, compute running aggregates (average fare, trip count per zone), and push updated metrics to a dashboard endpoint. Stale metrics older than 5 minutes are worse than no metrics.

Pipeline

[ry_ingest_events]
     |
     v
[ry_process_stream]
     |
     v
[ry_update_aggregates]
     |
     v
[ry_check_alerts]

Workflow inputs: eventBatch, windowSize, alertRules

Workers

CheckAlertsWorker (task: ry_check_alerts)

Checks alert rules against current aggregates and triggers alerts.

• Parses strings to int
• Writes alerts, alertCount

IngestEventsWorker (task: ry_ingest_events)

Ingests a batch of events for real-time analytics processing.

• Writes events, eventCount

ProcessStreamWorker (task: ry_process_stream)

Processes the event stream: computes window metrics, flags anomalies.

• Writes processed, processedCount, windowMetrics

UpdateAggregatesWorker (task: ry_update_aggregates)

Updates running aggregates with the latest window metrics.

• Writes aggregates, updatedCount

---

20 tests | Workflow: real_time_analytics | Timeout: 60s

See RUNNING.md for setup and usage.