Worked on enhancing memory monitoring within the starkware-libs/sequencer repository by upgrading pod memory pressure detection to use container_memory_rss metrics. This approach replaced the previous working set metric, resulting in more accurate and actionable alerts for Kubernetes environments. The update included revising dashboard panels and documentation to reflect the new RSS-based thresholds, improving both alert accuracy and resource planning. Leveraged Rust for backend development and focused on monitoring and alerting best practices. Quality was maintained through retriggered CI checks and collaborative code reviews, ensuring that the changes reduced alert fatigue and provided clearer guidance for operational teams.