Worked on the opensearch-project/index-management repository to deliver a feature that changes snapshot creation behavior, ensuring global state is excluded by default and only index data is captured. This adjustment improves the isolation and predictability of snapshot operations, reducing the risk of unintentionally including cluster-wide settings during backups and restores. The implementation focused on backend development using Kotlin, aligning with OpenSearch repository standards and practices. Through targeted commits and collaborative pull request reviews, the work enhanced reliability and simplified compliance for snapshot data management, demonstrating a methodical approach to feature delivery and a clear understanding of operational requirements in distributed systems.