Software Engineer, Systems ML

Meta is seeking a Software Engineer to join our Systems ML Engineering team, focused on building and optimizing the machine learning infrastructure that powers Meta's products at massive scale. In this role, you will design and develop high-performance ML systems, working across the full stack from model training and inference pipelines to hardware-aware optimizations. You will collaborate with researchers, platform engineers, and product teams to accelerate ML workloads and improve the efficiency of AI infrastructure that serves billions of users.

Responsibilities

Design, build, and optimize large-scale ML training and inference systems, including distributed computing frameworks and hardware-accelerated pipelines
Develop and maintain high-performance ML infrastructure components in C++ and Python, ensuring reliability, scalability, and low-latency execution
Identify and resolve performance bottlenecks across the ML stack using profiling, instrumentation, and benchmarking tools
Architect and evaluate trade-offs in ML system design, including memory bandwidth, compute utilization, and I/O throughput
Partner with research and product teams to translate ML model requirements into efficient infrastructure solutions
Write automated tests covering expected behaviors, failure modes, and error paths for ML infrastructure components, and build monitoring and alerting for production anomalies
Contribute to staged rollout strategies using feature flagging and experimentation frameworks to safely deploy ML system changes
Produce accurate technical documentation when introducing new ML infrastructure features or updating existing systems
Participate in on-call rotations to investigate and mitigate production incidents affecting ML serving systems, and contribute detailed retrospectives
Provide constructive code review feedback to other engineers and contribute to engineering programs that improve the health and quality of the ML infrastructure codebase

Qualifications

Currently has, or is in the process of obtaining a Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience. Degree must be completed prior to joining Meta
2+ years of experience in software engineering with a focus on machine learning systems, AI infrastructure, or high-performance computing
Experience developing or optimizing ML training or inference pipelines using frameworks such as PyTorch, TensorFlow, or equivalent
Experience with distributed computing architectures and large-scale systems design for ML workloads
Experience programming in C++ and Python for performance-critical systems
Experience using profiling and performance analysis tools to identify and resolve bottlenecks in ML or compute-intensive systems 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)
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 with ML compiler technologies such as MLIR, LLVM, TVM, XLA, or IREE
Experience with GPU programming using CUDA, ROCm, or equivalent hardware accelerator kernel development
Experience adhering to and implementing responsible, ethical AI practices (e.g., risk assessment, bias mitigation, quality and accuracy reviews)
Demonstrated ongoing AI skill development (e.g., prompt/context engineering, agent orchestration) and staying current with emerging AI technologies
Experience optimizing large-scale ranking or recommendation model inference on AI accelerator hardware such as GPUs or TPUs
Experience with hardware-software co-design, including numerics optimization and SIMD or vectorization techniques