A finance company's customers request appointments with advisors via text messages. A web application on Amazon EC2 accepts requests and publishes the text messages to an Amazon Simple Queue Service (Amazon SQS) queue. Another EC2 application sends meeting invitations and confirmation emails, then stores meeting details in an Amazon DynamoDB table. As the company grows, customers report meeting invitations are taking longer to arrive. What should a solutions architect recommend to resolve this?

The correct solution addresses the bottleneck in sending meeting invitations. As customer requests increase, the application responsible for sending invitations and emails becomes overloaded, leading to delays. An Auto Scaling group for this application, configured to scale based on the SQS queue depth, ensures that more instances are launched automatically when the queue backlog grows. This allows the system to process messages faster and reduce delays in sending invitations. Adding a DAX cluster would improve read performance for DynamoDB, but the problem is with processing new requests, not reading existing data. Adding API Gateway or CloudFront would help with the initial web application's scalability and performance, but the core issue is the downstream processing of messages from the SQS queue.