Worked on the kyverno/kyverno repository to enhance backend stability by refactoring the leader election process using Go. Addressed a critical bug by modifying the startup flow so that errors are now returned instead of causing the application to terminate abruptly, which allows for graceful failure handling and improved observability in distributed environments. This change improved startup reliability and reduced downtime by enabling explicit error propagation. The work focused on backend development, emphasizing maintainability and future resilience through clearer error handling paths. The contribution did not introduce new features but delivered a targeted fix that strengthened the robustness of the deployment process.