Worked on improving the reliability of metrics in the maticnetwork/bor repository, focusing on backend development and metrics monitoring using Go. Addressed a bug in the ResettingSample component where Prometheus _count metrics were not monotonic, which could lead to inaccurate dashboards and unreliable alerting. Introduced an atomic counter to ensure correct metric increments and implemented a Clear method to allow deterministic resetting of samples and state. Expanded test coverage to validate both monotonicity and reset behavior, reducing the risk of regression. These changes enhanced the accuracy of operational visibility and improved the reliability of monitoring across services using Bor metrics.