Worked on backend reliability and network resilience for UniversalMediaServer and redis/lettuce, focusing on concurrency, error handling, and maintainability. In UniversalMediaServer, introduced a shared DNS resolver with configurable timeouts, refactored networking components for improved logging and error handling, and expanded test coverage using Java and asynchronous programming. For redis/lettuce, addressed a deadlock in cluster topology refresh by implementing robust error handling for invalid URIs and adding targeted unit tests to prevent regressions. These contributions enhanced system stability, observability, and developer workflow, demonstrating a methodical approach to backend development and a strong emphasis on testability and long-term maintainability.