Event-Driven Architecture

In event-driven systems, services publish events when something changes and other services react by consuming those events. This decouples producers and consumers and enables near real-time workflows.

When it fits

• High-volume, real-time updates or workflows
• You want to minimise direct dependencies between services

Risks

• Event ordering, idempotency and debugging distributed flows
• Event schema evolution and replay strategies

Real-world examples

• Ride-sharing platforms stream trip, location and payment events through Kafka or Kinesis to power pricing and ETAs
• E-commerce platforms publish order and inventory events to update fulfilment, notifications and analytics in near real time

Practical tips

• Define clear event contracts and version them
• Make consumers idempotent and design for at-least-once delivery