AWS, Azure & GCP · Cloud across providers
How do event-driven architectures work?
Short answer
Services emit events when something happens, and other services react asynchronously through brokers, streams or rules. This reduces coupling and enables fan-out, but requires idempotency, schema evolution and tracing.
An event describes something that happened, such as an order being created. The producer does not know or care who consumes it. A broker or router delivers it to interested consumers, which may number zero or many.
Patterns
- Pub/sub for fan-out to multiple consumers, such as SNS or Pub/Sub
- Event bus with rules and routing, such as EventBridge or Event Grid
- Streaming log for ordered replay, such as Kafka or Kinesis
- Choreography where services react to each other, versus orchestration with a central workflow
{
"id": "evt_123",
"type": "order.created",
"source": "orders",
"time": "2026-10-28T10:00:00Z",
"data": {
"orderId": "ord_456",
"total": 2499
}
}Because delivery is asynchronous, consumers must tolerate duplicates and out-of-order events. Include an event id for deduplication and a version in the schema. Prefer additive changes, and keep a contract test so producers do not break consumers silently.
Follow-ups you should expect
Interviewers use these to check whether the first answer was understanding or memorisation.
- 1What is the difference between an event and a command?
- 2How do you handle duplicate or out-of-order events?
- 3When would you use orchestration instead of choreography?
- 4How do you evolve an event schema safely?
Say it in the interview
In an event-driven architecture, a service publishes an event describing what happened, and consumers react asynchronously through a broker or stream. That reduces coupling and allows fan-out, but I have to design for duplicate delivery, ordering and schema evolution. I include an event id and version, make consumers idempotent, and trace the whole flow because debugging is harder than a synchronous call.
Use this as a model, not a script. Saying it in your own words is what makes it survive follow-up pressure.
Related questions
Reading is step one
Now say it out loud.
Understanding an answer and delivering it under pressure are different skills. Rehearse this topic in a live AWS, Azure & GCP mock interview and get scored feedback on depth, clarity and confidence.