A company uses an EventBridge rule that targets a Lambda function to auto-remediate non-compliant S3 buckets detected by AWS Config. During an incident review, the team discovers that several remediation events silently failed and were never retried, leaving buckets publicly accessible for hours. Investigation shows the Lambda function was throttled due to a spike in events, and EventBridge exhausted its retries. The team needs to ensure that no remediation event is ever lost and that failed events can be reprocessed after the throttling condition clears. What should the DevOps engineer implement?