Event-Driven Order Pipeline

Event-driven architecture is the foundation of modern distributed systems, enabling loose coupling, independent scaling, and resilient processing. In this challenge, you will design an order processing pipeline for an e-commerce platform that handles thousands of orders per minute with exactly-once processing guarantees and full audit trails. The architecture starts with an API Gateway endpoint accepting order submissions, which publishes order events to Amazon EventBridge. EventBridge rules route events to different processing pipelines: payment processing, inventory reservation, shipping label generation, and notification sending. Each pipeline runs as an independent Step Functions state machine, providing visual workflow tracking, automatic retries with exponential backoff, and compensation logic for failures. The payment pipeline integrates with a payment gateway via Lambda, implementing the saga pattern for distributed transactions — if payment succeeds but inventory reservation fails, a compensating transaction reverses the payment. Inventory management uses DynamoDB with conditional writes to prevent overselling, using atomic counters and condition expressions for race-condition-free stock management. The event store pattern captures every state change as an immutable event in DynamoDB Streams, enabling full order history reconstruction and audit compliance. For analytics, DynamoDB Streams feed into Kinesis Data Firehose, which delivers to S3 in Parquet format for Athena queries. Dead-letter queues capture failed events at every stage, with a Lambda function that attempts automated remediation before alerting operations. The architecture includes an order status API using DynamoDB queries with GSI on customer ID and order status. This challenge teaches event-driven architecture, the saga pattern for distributed transactions, and event sourcing fundamentals.