Member of Technical Staff, ML

Job Title

Member of Technical Staff, ML

Role Summary

Work on training and optimizing large language models to run efficiently on MatX hardware. Contribute to distributed training and inference infrastructure, model evaluation, and ML-driven hardware co-design.

Experience Level

Mid-level. Candidates should have demonstrated experience in LLM research or production ML engineering and the ability to work across research and systems teams.

Responsibilities

Primary responsibilities include:

• Train, fine-tune, and optimize LLMs for MatX hardware.
• Design, build, and operate distributed training and inference infrastructure.
• Run and develop model quality evaluations and benchmarks.
• Advise hardware architecture from an ML performance and numerics perspective.
• Co-design hardware using ML experiments and RTL prototypes.
• Collaborate with cross-functional teams to validate numerics and productionize models.

Requirements

Must-have qualifications and constraints:

• Strong research and engineering background in LLMs with depth in at least one area: reinforcement learning, quantization, or attention/architecture (including long-context methods).
• Proven software engineering skills building and scaling distributed training and inference codebases.
• Experience with inference/training optimization or ML–hardware co-design is preferred.
• Authorized to work in the United States and able to work from the Mountain View office Tuesday–Thursday.
• Must be able to comply with U.S. export control requirements for controlled information.

Nice-to-have:

• Publications or shipped open-source research in LLMs.
• Experience with RTL prototypes, hardware-aware numerics, or large-scale evaluation systems.

Education Requirements

PhD in Machine Learning, Computer Science, Electrical Engineering, or a related field, or equivalent practical experience; demonstrated research track record in LLMs via publications or shipped/open-source work is acceptable.

About the Company

Company: MatX

Headquarters: Mountain View, California, USA

MatX specializes in creating faster chips for large language models (LLMs), focusing on innovative hardware and software solutions. The company fosters a collaborative and supportive work environment, welcoming candidates of all experience levels. Their approach prioritizes deep understanding and consideration of novel methods to drive efficiency and performance in their projects, particularly in silicon design and related engineering roles.