Worked on enhancing the valkey-io/valkey-glide repository by developing a liveness-aware pipeline send-timeout mechanism to improve reliability under network backpressure. Leveraging Rust and asynchronous programming, the solution distinguished between slow-but-live and dead connections, reducing unnecessary command failures and ensuring consistent command delivery. The implementation included a configurable buffer sizing option and comprehensive test coverage, validating behavior across live, slow, and terminated connection scenarios. Design documentation and changelogs were updated to clarify root-cause analysis and configuration guidance. Code quality was maintained through formatting, documentation improvements, and collaborative reviews, resulting in more stable and observable backend deployments for production environments.