Software Engineer, Systems ML

Meta is seeking a Research Engineer specializing in Systems Machine Learning to help design and build the infrastructure and algorithmic foundations that power large-scale AI systems across Meta's product ecosystem. In this role, you will work at the intersection of machine learning research and systems engineering, developing novel approaches to training efficiency, model serving, distributed computation, and hardware-software co-design. You will collaborate with research scientists and product engineers to translate cutting-edge ML research into production-grade systems that operate at massive scale, directly shaping the performance and reliability of Meta's AI-driven products. Design and implement scalable systems for distributed ML training and inference, including optimizations across compute, memory, and communication bottlenecks Develop and evaluate novel techniques for accelerating AI research workflows such as training, inference, RL, evals on latest generation hardware platforms Lead the architecture and end-to-end delivery of major systems ML initiatives, coordinating across research scientists, product engineers, and external partners Establish performance benchmarking frameworks and profiling pipelines to identify bottlenecks and drive measurable improvements in training throughput and inference latency Define service level objectives and reliability standards for ML training and serving systems, building dashboards and runbooks to reduce incident response time Apply AI-assisted development workflows to accelerate implementation, code review, and systems analysis, serving as a model for AI-native engineering practices within the team Collaborate with cross-functional partners in infrastructure, and product engineering to co-design ML systems that maximize research velocity and researcher experience Mentor other engineers on systems ML best practices, distributed training patterns, and debugging methodologies for large-scale ML infrastructure Communicate technical trade-offs, architectural decisions, and experimental results clearly to both engineering and research audiences through design documents and presentations Contribute to the broader research community by publishing findings on systems ML advances at leading venues Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience Bachelor's degree in Computer Science, Electrical Engineering, or a related technical field 8+ years of experience in systems engineering, machine learning infrastructure, or a closely related field Experience designing and optimizing distributed ML training or inference systems at scale, including proficiency with frameworks such as PyTorch, JAX, or TensorFlow Experience with low-level systems programming in C++ or CUDA, including performance profiling, kernel optimization, or compiler-level ML optimizations Experience leading the technical design and delivery of complex, cross-functional systems ML projects from inception through production deployment Experience using data-driven methods and experimentation to evaluate and validate systems performance improvements Master's or PhD degree in Computer Science, Electrical Engineering, Machine Learning, or a related technical field Track record of publishing research on systems ML topics at venues such as MLSys, OSDI, SOSP, NeurIPS, or ICML Demonstrated ongoing AI skill development (e.g., prompt/context engineering, agent orchestration) and staying current with emerging AI technologies Demonstrated ability to integrate AI tools to optimize/redesign workflows and drive measurable impact (e.g., efficiency gains, quality improvements) Experience adhering to and implementing responsible, ethical AI practices (e.g., risk assessment, bias mitigation, quality and accuracy reviews) Experience with ML compiler stacks such as MLIR, XLA, TVM, or Triton, and familiarity with hardware-software co-design for AI accelerators Experience building automated tooling or frameworks that improve engineering efficiency across ML infrastructure teams Experience with model parallelism strategies including tensor parallelism, pipeline parallelism, and expert parallelism for large-scale model training