Worked on the cozystack/cozystack repository to enhance monitoring and alerting for Kubernetes environments, focusing on LINSTOR integration. Delivered a composite feature that separated availability and metrics alerts, improved alert labeling, and introduced a hold-time mechanism to reduce alert flooding, all implemented using YAML and leveraging configuration management best practices. Addressed a critical bug by correcting VMPodScrape job labels, ensuring accurate metrics collection and monitoring from the intended Kubernetes components. These changes improved observability for operators, supported faster mean time to resolution, and enabled more reliable capacity planning, demonstrating a strong grasp of alert management, DevOps workflows, and monitoring systems.