A developer must transfer expired items from a DynamoDB table to Amazon S3. The table uses TTL to expire items. The application must process each expired item and then store it in S3; processing plus storage takes 5 minutes per item. Which solution accomplishes this with the least operational overhead?

DynamoDB Streams is the most efficient and least operationally intensive solution for processing expired items. When an item expires due to TTL, DynamoDB Streams captures this event as a REMOVE record. A Lambda function triggered by this stream can then process each expired item record and store it in Amazon S3. This approach is serverless, scales automatically, and only incurs costs when items expire and are processed, aligning perfectly with the requirement for minimal operational overhead. DAX is a caching service for DynamoDB and does not help with processing expired items. A custom application on ECS would require managing servers and scaling, increasing operational overhead. An EventBridge rule could trigger a Lambda, but it would need to periodically scan the table for expired items, which is less efficient and more costly than reacting directly to TTL events via DynamoDB Streams.