The correct answer is: Send an image upload message to Amazon SQS for downstream thumbnail processing. Immediately notify the user in the application that the upload was successful.

This architecture enables asynchronous decoupling between the image upload tier and the processing tier. When an image is uploaded, the application stores it (e.g., in S3), places a message on an Amazon SQS queue, and returns a success message to the user right awayachieving the goal of minimal user-facing latency. The backend processing tier can poll the SQS queue and process thumbnail generation independently of the upload workflow, enabling scalability and fault-tolerant retries.

The option Use Amazon SNS to create two topics: one for launching thumbnail processing after upload and another for sending a push notification once processing completes adds complexity and does not provide durability or retry capabilities as effectively as SQS. SNS is better for broadcast-style fan-out notifications, not queued asynchronous job processing.

The option Deploy a Lambda function triggered by the image upload event to both generate the thumbnail and notify the user upon completion is not optimal due to the 60-second processing time. AWS Lambda has timeout limits (default of 3 seconds, max of 15 minutes), and using it synchronously risks a timeout or delays in acknowledgment to the user.

The option Build an AWS Step Functions workflow to coordinate upload, thumbnail generation, and user confirmation in a sequential state machine unnecessarily serializes the process. It delays user notification until all processing steps are complete, which defeats the goal of immediate acknowledgment. Step Functions is best for orchestrating longer workflows that require visibility, retries, or branchingnot for fast, decoupled user interactions.

This question is important because it evaluates your ability to design scalable, user-responsive, decoupled architectures using AWS messaging services. SQS is purpose-built for job queueing and background task handling, especially when paired with scalable consumers like EC2 or Lambda.

Tip: When you see requirements for immediate user feedback and long-running background processing, think of queuing mechanisms like SQS. If the problem is event fan-out or notification-based, SNS is likely the answer. For multi-step workflows, Step Functions is appropriatebut not when user interaction must be decoupled from processing speed.

Full AWS Practitioner Certification Question