Worked on enhancing server observability and reliability for the opendatahub-io/feast and red-hat-data-services/feast repositories, focusing on backend development using Go. Delivered features that introduced comprehensive metrics and tracing for both HTTP and gRPC servers, leveraging Prometheus to monitor HTTP requests and improve operational visibility. Hardened server shutdown logic to prevent deadlocks, reducing deployment risks and ensuring safer restarts. Updated documentation to clearly outline new metrics, tracing capabilities, and shutdown procedures, supporting better troubleshooting and performance monitoring. These improvements enabled more data-driven capacity planning and faster incident response, reflecting a methodical approach to backend reliability and observability engineering.