An e-commerce application receives and processes orders, then stores order data in Amazon DynamoDB. Traffic is sporadic but can spike during campaigns; the system must scale and process orders with minimal delay. Which design is the MOST reliable, simple, highly available, and loosely coupled way to receive and process orders?

Receiving orders into an Amazon SQS queue and invoking an AWS Lambda function to process them is the most reliable, simple, highly available, and loosely coupled solution. SQS provides a highly available, scalable, and durable buffer for incoming orders, decoupling the order receipt from processing. Lambda functions are serverless, scale automatically with demand, and are invoked asynchronously by SQS, ensuring orders are processed with minimal delay even during traffic spikes. This architecture eliminates the need to manage servers, simplifying operations. Using a database on EC2 for order receipt introduces a single point of failure and requires manual scaling. AWS Step Functions add unnecessary complexity for simple order processing. Kinesis Data Streams are suitable for real-time analytics but might be overkill for basic order processing and still require managing EC2 instances for consumers.