A company’s data scientists use Amazon SageMaker to read, process, and write data to an Amazon S3 bucket. Each project has a dedicated S3 prefix, and bucket policies restrict access by prefix. Projects last up to 60 days, and the security team requires that data not remain in S3 after a project ends. What is the MOST cost-effective solution?

The most cost-effective solution is to create an S3 Lifecycle rule for each project's prefix to expire objects after 60 days. S3 Lifecycle rules are a native S3 feature designed for automated object management, including expiration, and are free to configure. This directly addresses the requirement to delete data after 60 days. Creating an AWS Lambda function to identify and delete objects would incur Lambda execution costs and potentially S3 API request costs, making it less cost-effective than a native lifecycle rule. Creating a new S3 bucket and enabling S3 Intelligent-Tiering would involve copying data, incurring data transfer and storage costs in the new bucket, and Intelligent-Tiering does not automatically delete objects after a set period. Triggering a Lambda function with S3 Intelligent-Tiering archival events is not a direct mechanism for object deletion based on age and would still incur Lambda costs.