Compiler Performance Engineer

This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a Compiler Performance Engineer based in India. As a Compiler Performance Engineer, you will help optimize a production-grade Java just-in-time (JIT) compiler to generate faster, more efficient machine code. You will analyze Java benchmarks and real-world enterprise workloads to identify performance bottlenecks and trace them to their root causes within the compiler. Working across the compiler's optimization and code generation layers, you will implement improvements that deliver measurable performance gains. You will collaborate with a distributed team of compiler specialists and contribute to the evolution of a mature compiler built on LLVM. Your work will also include developing tests and benchmarks to validate improvements and prevent regressions. This role offers an opportunity to apply low-level programming expertise, compiler theory, and hardware knowledge to solve challenging performance problems with direct customer impact.

Analyze Java benchmarks and customer workloads to identify performance gaps in generated machine code and determine their underlying causes within the compiler. Implement and refine compiler optimizations across the middle end and back end, including inlining, loop transformations, vectorization, redundancy elimination, and speculative optimizations. Investigate how compiler optimizations interact with runtime behavior and identify opportunities to improve execution efficiency. Improve machine code generation for x86_64 and AArch64 processor architectures. Evaluate optimization strategies and assess their impact on runtime performance, code quality, and execution efficiency. Develop and maintain tests and benchmarks that demonstrate performance improvements and protect against regressions. Use systematic performance analysis and debugging techniques to investigate complex compiler behavior and validate proposed solutions. Contribute relevant improvements to upstream LLVM, supporting the continued advancement of shared compiler infrastructure. Collaborate with compiler engineers in a distributed team to review solutions, share technical insights, and deliver measurable performance improvements. Requirements: At least 2 years of professional experience in compiler development, JIT compilers, or low-level performance engineering. Strong C++ programming and software design skills, including the ability to work effectively with complex codebases. Solid understanding of compiler optimization theory, including Static Single Assignment (SSA), data flow analysis, and loop transformations. Ability to read and interpret assembly code and explain why particular instruction sequences perform better than others. Strong analytical and problem-solving skills, with the ability to investigate performance bottlenecks and translate findings into effective technical solutions. Ability to develop reliable tests and benchmarks to validate optimizations and ensure performance improvements are repeatable. Effective collaboration and communication skills for working with a distributed team of engineering specialists. Preferred qualifications: Experience with LLVM Intermediate Representation (IR) and LLVM optimization passes, or another optimizing compiler such as HotSpot C2, Graal, GCC, or V8. Familiarity with performance profiling tools such as perf , VTune, or async-profiler. Knowledge of x86_64 and/or AArch64 microarchitecture and its impact on generated code performance. Experience with Java, Scala, Kotlin, or JVM internals. A master's degree or PhD in Computer Science, particularly with a focus on compilers or performance engineering. Benefits: Remote work opportunity within India. Opportunity to work on a production-grade Java JIT compiler used in enterprise environments. Hands-on experience solving challenging compiler optimization and low-level performance problems. Collaboration with a distributed team of compiler experts. Opportunity to contribute to LLVM and other compiler optimization initiatives. Direct, measurable impact on application performance across real-world workloads. Opportunities to deepen expertise in C++, compiler architecture, machine code generation, and modern processor architectures.