A company is building a web application on AWS. When a customer requests a report, the application will generate the report and make it available to the customer within one hour. Reports should remain accessible for 8 hours. Some reports exceed 1 MB. Each report is unique to the requesting customer. The application must delete reports older than 2 days. Which approach provides the required behavior with the LEAST operational overhead?

Storing reports in Amazon S3 is ideal for large, unique files like these reports, as S3 is highly scalable and cost-effective for object storage. Server-side encryption ensures data security. S3 presigned URLs provide secure, temporary access to specific objects, fulfilling the 8-hour accessibility requirement without needing to manage user authentication for S3 directly. S3 Lifecycle rules automate the deletion of objects older than 2 days, minimizing operational overhead. DynamoDB is not suitable for storing large reports (exceeding 1MB) as it's a NoSQL database optimized for key-value and document data, not large binary objects. Attaching reports to SNS messages is impractical because SNS messages have a size limit (256 KB), making it unsuitable for reports exceeding 1 MB. Storing reports in an RDS database is less efficient and more expensive for large binary files compared to S3, and managing deletions with a Lambda function adds more operational overhead than S3 Lifecycle rules.