Member of Technical Staff, Inference Systems Research (AI)

Location: Mountain View, United States. Employees living within 50 miles of the designated U.S. office are expected to work from the office at least four days per week.

Salary: USD $188,000–$304,200 per year for the San Francisco Bay Area and New York City metropolitan area; a U.S.-wide range of USD $142,800–$274,800 also applies to other locations.

Company

Microsoft AI is a startup-like organization within Microsoft focused on advancing safe and responsible artificial intelligence and frontier models.

What you will do

• Work with researchers and engineers to implement frontier AI research ideas.
• Develop systems, tools, and techniques that improve model inference performance.
• Build debugging tools for performance bottlenecks, numerical instabilities, and distributed systems issues.
• Create tools and processes that improve collective engineering productivity.
• Remove technical roadblocks and deliver improvements to users quickly and iteratively.

Requirements

• Bachelor's degree in Computer Science or a related technical field and 6+ years of technical engineering experience, or equivalent experience.
• Experience with generative AI, distributed computing, and large-scale production inference.
• Expertise in Python and its ecosystem, including tools such as uv, pybind/nanobind, and FastAPI.
• Experience with GPU kernel programming, benchmarking, profiling, and optimizing PyTorch generative AI models.
• Familiarity with open-source inference frameworks such as vLLM and SGLang, as well as the JAX scaling book.
• Working knowledge of programming languages including C, C++, C#, Java, JavaScript, or Python.

Nice to have

• Master's degree in Computer Science or a related technical field and 8+ years of technical engineering experience, or a bachelor's degree and 12+ years of experience.
• Experience working in a fast-paced, design-driven product development cycle.

Culture & Benefits

• Collaboration with researchers in Microsoft AI's research organization.
• Work across kernels, inference algorithms, model architecture co-design, ASIC co-design, distributed systems, and profiling tools.
• Focus on respect, integrity, accountability, inclusion, and a growth mindset.
• Potential eligibility for benefits and additional compensation.