Tooling Engineer

MATS Research • Berkeley, CA

45d • $180,000 - $275,000 • Hybrid

About The Position

MATS is hiring a Tooling Engineer to build tools and infrastructure that make research compute and research workflows easier to use, operate, and scale. The role supports compute access across GPU providers, API providers, cloud platforms, and MATS’s own GPU cluster. Work includes automation and interfaces for provisioning, budgets, usage, and requests; unified access to compute; and controls and observability around cost, access, and usage.

The position also involves adopting, extending, or building agent, evaluation, and data infrastructure for AI safety research. The engineer will work closely with the Compute team, other MATS teams, fellows, residents, and MATS’s security function to identify friction and deliver useful, reusable solutions. Work may span internal tooling, automation, research tooling, and contributions to GPU cluster and cloud infrastructure, with priorities evolving as needs change. The role offers substantial ownership, while favoring adoption and extension of existing tools when they fit.

The engineer is expected to maintain, document, and make their work deployable by others on the team, and to iterate with real users. The position is 40 hours per week and is primarily in person at MATS’s main office in Berkeley, California; hybrid arrangements are available for exceptional candidates, but fully remote work is not offered.

Requirements

• At least 3 years of professional software engineering experience, including production code shipped and maintained.
• Strong software engineering skills, including proficiency in Python and the ability or willingness to become effective across a web stack.
• Experience integrating third-party APIs and building automations or data pipelines against them.
• Comfort working directly with users, gathering requirements iteratively, shipping small, and refining based on real usage.
• Effective use of modern AI coding tools and agents, such as Claude Code or Cursor, as part of the software-development workflow.
• Clear written communication and the ability to document work.
• High agency and ability to take an ambiguous problem, scope it, and drive it to a solution people use.
• Alignment with MATS’s mission of reducing risk from the development of advanced AI.

Nice To Haves

• Experience building internal tools, dashboards, or platforms for researchers, engineers, or other technical users.
• Experience with research or ML tooling, such as agent infrastructure, evaluation systems, experiment tooling, research-data pipelines, or other researcher-facing engineering.
• Experience with React and TypeScript.
• Cloud administration and cost control on GCP or AWS; familiarity with GPU providers, inference routers, and how researchers use them.
• Data engineering experience, including ETL, analytics storage, and reconciliation or metrics pipelines.
• Linux, HPC, or ML infrastructure experience, such as Slurm clusters; useful but not required.
• Prior work in an AI safety organization, fellowship, or research lab.

Responsibilities

• Build tools and automation for compute access, provisioning, budgets, usage, and requests, and automate recurring workflows across MATS.
• Work with fellows and the Compute team to understand workflows, identify operational bottlenecks, and develop reusable tools.
• Build simpler, more unified ways to access and manage compute across supported GPU, API, and cloud providers, adopting or extending existing tools where appropriate.
• Contribute where useful to GPU cluster and cloud infrastructure, including controls and observability for cost, access, and usage.
• Work with MATS’s security function on security and access requirements for research infrastructure and tooling.
• Build reliable pipelines and interfaces for compute usage, spending, capacity, and operational state, with clear provenance and strong data-quality standards.
• Make compute and operational information useful to fellows, the Compute team, and organizational planning and decision-making.
• Adopt, extend, or build AI safety research tools, including agent and evaluation infrastructure and research data infrastructure.
• Work with fellows and residents to understand research workflows and build reusable tools that accelerate research across projects.
• Keep tools maintainable, documented, and deployable by others on the team, and iterate with real users.

Benefits

• In-office catered lunches and dinners on workdays.
• Paid work trips, including staff retreats, business trips, and relevant conferences.
• Funding and support for professional development.
• 17 days of annual paid time off for Berkeley-based employees.
• Flexibility for hybrid work for exceptional candidates; fully remote work is not available.
• Collaborative and intellectually stimulating work environment.
• Medical, dental, vision, and life insurance.
• Roth 401(k) and Traditional 401(k) for US-based employees.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed