EVENT-TIME VIEW
Live business activity
Orders Captured amount
Watermark Results through 14:03 are considered complete
REAL-TIME ORDER INTELLIGENCE
Overview
BUSINESS OUTCOME
What is happening right now
LIVE EVIDENCE
Recent outcomes
OPERATIONAL VIEW
Orders and payments
| Order | Customer | Order status | Amount | Payment | Event time | Latency |
|---|
MOCK INPUT
Produce a simulated Kafka event
Enter an order or payment and watch the mock dashboard show the same flow the local implementation proves with Kafka, Flink, and PostgreSQL.
LIVE KAFKA EVIDENCE
Show the message moving through Kafka
- In the local implementation, publish the event above.
- Open Kafka UI and select local-orderflow.
- Open Topics, then orders.v1.
- Use the Messages tab and search for the returned event ID.
TRUSTED RESULTS
Data quality evidence
REJECTION AUDIT
Invalid events
TECHNICAL PROOF
Pipeline health
CONTROLLED DEMONSTRATION
Demo Lab
Run one observable test at a time. Every scenario includes its expected result and verification.
SYSTEM DESIGN
How the platform works
01
INGEST
Kafka
Durably stores replayable order and payment events while separating producers from stream processing.
- orders.v1
- payments.v1
- events.dlq.v1
02
PROCESS
Apache Flink
Validates, deduplicates, correlates, and aggregates events using event time and managed state.
- 10s watermarks
- 24h deduplication TTL
- Checkpoint recovery
03
SERVE
PostgreSQL
Exposes current order state, minute-level metrics, and rejection records for operational use.
- order_current
- order_metrics_1m
- rejected_event