In August 2025, Bilal Khan contributed to the apache/auron repository by enhancing Spark’s dynamic class loading reliability. He addressed a ByteBuddy integration issue by switching to the contextClassLoader for dynamic redefinitions, improving the robustness of modules like ForceApplyShuffledHashJoinInjector and ValidateSparkPlanInjector in distributed Spark environments. Additionally, Bilal refactored summation logic across critical data-path modules, replacing map and sum patterns with foldLeft to streamline code and potentially boost performance. His work, primarily in Java and Scala, focused on maintainability and efficiency, demonstrating a solid understanding of data engineering challenges and functional programming within large-scale Spark systems.