As a Software Engineering Manager in the MTIA Software team, you will lead the development of Meta's ML compiler stack for custom AI accelerators - the layer that enables kernels, both hand-written and generated, run efficiently on MTIA silicon. Your team defines how Triton and lower-level DSLs map onto MTIA hardware, shaping language extensions, target-specific optimizations, and the kernel authoring experience used by kernel engineers, product groups, and model teams across Meta. The work spans open-source contributions and internal compiler development, directly influences ISA and architecture roadmap decisions, and carries top-level executive visibility as a key AI infrastructure initiative. The team culture celebrates deep technical contributions with extensive knowledge-sharing, including external conference talks and published research.
Responsibilities
You will drive the development of the DSL compiler toolchain, including ML compiler integration, GEMM and non-GEMM kernel enablement, collectives support, and next-generation low-level DSLs. You will lead teams of engineers and technical leaders delivering large-scale projects across compiler infrastructure, kernel libraries, and production deployment on multiple accelerator generations
• You will stay technically engaged, contributing hands-on through AI tools to evaluate the technical direction and quality across your teams. You will form trusted cross-functional partnerships with hardware/architecture, kernel, runtime, framework, and model teams while actively shaping strategy and roadmap, and you will balance near-term deployment commitments against long-term compiler investments. You will build and mentor engineers into AI-era leaders while maintaining high engineering craft standards
Minimum Qualifications
• 5+ years of experience in managing a software team in a fast-paced capacity
• Experience with compiler architecture and development, particularly with ML compilers or domain-specific languages (DSLs)
• Proven understanding and experience in executing full product life-cycles (prototyping, deployment, and support)
Preferred Qualifications
• Experience working closely with hardware architectures such as SIMD, GPU, RISC-V, AI accelerators, and influencing architecture roadmaps
• Experience with Triton, MLIR, or LLVM compiler infrastructure
• Experience with open-source community engagement and upstream contributions
• Demonstrated ability to integrate AI tools to optimize/redesign workflows and drive measurable impact (e.g., efficiency gains, quality improvements)
• Demonstrated ability to integrate AI tools to optimize/redesign workflows and drive measurable impact (e.g., efficiency gains, quality improvements)
• Experience with PyTorch compiler stack internals such as Inductor
• Demonstrated ongoing AI skill development (e.g., prompt/context engineering, agent orchestration) and staying current with emerging AI technologies
• Demonstrated ongoing AI skill development (e.g., prompt/context engineering, agent orchestration) and staying current with emerging AI technologies
• Experience with different programming models for high-performance computations, e.g., GPU CUDA programming or OpenCL or OpenMP programming
• Experience adhering to and implementing responsible, ethical AI practices (e.g., risk assessment, bias mitigation, quality and accuracy reviews)
• Experience adhering to and implementing responsible, ethical AI practices (e.g., risk assessment, bias mitigation, quality and accuracy reviews)
• Experience with kernel-level performance optimization for AI accelerators
