About the job
Join the Systems Planning and Architecture (SPARC) team within Microsoft’s Azure Hardware Systems and Infrastructure (AHSI) organization, the team behind Microsoft’s expanding Cloud Infrastructure and for powering Microsoft’s “Intelligent Cloud” mission. We are seeking a highly skilled Senior AI Software Architect to join our team focused on model enablement and performance optimization for Maia accelerators. This role is ideal for someone with strong experience in PyTorch-based model development, quantization techniques, and parallelization strategies at the framework level. You will work closely with hardware and software teams to bring up models on Maia and ensure they run efficiently.
Responsibilities
Port and optimize large-scale AI models (e.g., foundation models, diffusion models, YOLO) to run efficiently on Maia hardware.
Integrate models using frameworks such as PyTorch, ONNX, vLLM, and SGLang.
Apply techniques like KV cache quantization (e.g., BF16 → FP8), checkpointing, and re-sharding for efficient inference and training.
Experiment with parallelism strategies (TP, PP) and analyze performance impacts across interconnects (NVLink vs PCIe).
Collaborate on improving inference pipelines, including KV caching in sglang/vllm and performance tuning at the PyTorch level.
Work with Triton kernels for basic operations (e.g., FP8 dequantization) and assist in kernel performance analysis.
Partner with hardware architects and kernel developers for co-design discussions.
Communicate effectively with multiple stakeholders to align on performance goals and deliverables.
Qualifications
Minimum
Bachelor's Degree in Computer Science or related technical field AND 4+ years technical engineering experience with coding in languages including, but not limited to, C, C++, C#, Java, JavaScript, or Python OR equivalent experience.
Ability to meet Microsoft, customer and/or government security screening requirements are required for this role. These requirements include but are not limited to the following specialized security screenings: Microsoft Cloud Background Check: This position will be required to pass the Microsoft Cloud Background Check upon hire/transfer and every two years thereafter.
Preferred
Bachelor's Degree in Computer Science or Engineering. 3+ years of strong hands-on experience with PyTorch and model optimization techniques. Practical knowledge of quantization techniques like PTQ/QAT especially for KV cache quantization. Familiarity with parallelization strategies and distributed training concepts (e.g., sharding, allreduce). 2+ years of experience with AI inference stacks like SGLang/vLLM and performance profiling. Excellent problem-solving and communication skills; ability to work in a collaborative team environment. 3+ years of experience in Triton kernels and CUDA programming (basic understanding is acceptable but willingness to learn is essential). Experience with AI accelerator hardware and embedded systems. 3+ years of prior work on efficient model checkpointing, resharding scripts, and large-scale model deployments for serving at scale.