AWS Certified Solutions Architect – Associate (SAA-C03) — Question 952

A company tracks customer satisfaction by using surveys that the company hosts on its website. The surveys sometimes reach thousands of customers every hour. Survey results are currently sent in email messages to the company so company employees can manually review results and assess customer sentiment.

The company wants to automate the customer survey process. Survey results must be available for the previous 12 months.

Which solution will meet these requirements in the MOST scalable way?

Answer options

• A. Send the survey results data to an Amazon API Gateway endpoint that is connected to an Amazon Simple Queue Service (Amazon SQS) queue. Create an AWS Lambda function to poll the SQS queue, call Amazon Comprehend for sentiment analysis, and save the results to an Amazon DynamoDB table. Set the TTL for all records to 365 days in the future.
• B. Send the survey results data to an API that is running on an Amazon EC2 instance. Configure the API to store the survey results as a new record in an Amazon DynamoDB table, call Amazon Comprehend for sentiment analysis, and save the results in a second DynamoDB table. Set the TTL for all records to 365 days in the future.
• C. Write the survey results data to an Amazon S3 bucket. Use S3 Event Notifications to invoke an AWS Lambda function to read the data and call Amazon Rekognition for sentiment analysis. Store the sentiment analysis results in a second S3 bucket. Use S3 lifecycle policies on each bucket to expire objects after 365 days.
• D. Send the survey results data to an Amazon API Gateway endpoint that is connected to an Amazon Simple Queue Service (Amazon SQS) queue. Configure the SQS queue to invoke an AWS Lambda function that calls Amazon Lex for sentiment analysis and saves the results to an Amazon DynamoDB table. Set the TTL for all records to 365 days in the future.

Correct answer: A

Explanation

Option A provides a highly scalable, serverless solution by utilizing Amazon API Gateway and Amazon SQS to decouple the ingestion tier and handle sudden traffic spikes gracefully. Amazon Comprehend is the correct service for natural language text sentiment analysis, and DynamoDB TTL automatically manages the 12-month data expiration requirement at no extra cost. Other options are incorrect because EC2 (Option B) is less scalable and requires management overhead, Amazon Rekognition (Option C) is for image and video analysis rather than text, and Amazon Lex (Option D) is for building conversational chatbots rather than performing sentiment analysis on raw survey text.