Worked on enhancing the reliability and observability of the DynamoDB Master Monitor within the Netflix/mantis repository. Focused on backend development using Java, the work addressed a bug by implementing robust exception handling to ensure that scheduling processes continued uninterrupted even when errors occurred. Exception metrics were introduced to improve system observability and facilitate more effective alerting. To validate these changes, targeted unit tests were added, ensuring the new exception-handling logic performed as intended. This approach improved the resilience and maintainability of the monitoring workflow, reducing the risk of missed schedules and supporting ongoing operational stability for the system.