Worked on the redis/go-redis repository to address resource management in observability tooling by implementing cleanup for OpenTelemetry metrics during client shutdown. Using Go and OpenTelemetry instrumentation, introduced logic to unregister metrics collectors when the Redis client closes, updating configuration and metrics reporting to handle shutdown states and channel lifecycles. This approach prevents resource leaks and stale metrics, enhancing reliability for long-running services. The work focused on lifecycle management and metrics instrumentation, ensuring that observability components are properly disposed of during shutdown. This fix improved resource efficiency and observability accuracy in production environments where Redis and OpenTelemetry are used together.