An event platform stores leaderboard results in DynamoDB and retains them for 30 days, after which a scheduled job deletes old data. The table uses fixed write capacity, but during busy months the scheduled delete job causes write throttling. The developer needs a long-term solution that removes old leaderboard records and optimizes write throughput. Which approach satisfies this?

Enabling a TTL (Time-to-Live) attribute is the most efficient and cost-effective solution. DynamoDB automatically deletes expired items in the background without consuming any write capacity from your table. This prevents write throttling during peak periods and eliminates the need for a separate scheduled job. Using DynamoDB Streams or AWS Step Functions would involve custom logic and consume write capacity, potentially leading to throttling or increased costs. Increasing write capacity is a temporary fix that incurs higher costs and doesn't address the underlying inefficiency of a scheduled delete job.