Compute Administrator (Berkeley)

Applications will be reviewed on a rolling basis.

About The Organization

MATS Research aims to find and train talented individuals for what we see as the world’s most urgent and talent-constrained problem: reducing risks from advanced artificial intelligence. Our mission is to solve the talent bottleneck in AI safety and security through training, credentials, and community. We believe that ambitious researchers from a variety of backgrounds have the potential to contribute to the fields of AI alignment, interpretability, governance, and security research. Through our research fellowship, we aim to provide the mentorship, financial support, and community necessary to aid this transition. Please see our website for more details.

The Role

As a Compute Administrator, you will play a crucial role in empowering our scholars and mentors to conduct high-impact AI alignment, control, security, and governance research. This role is both systems-facing and researcher-facing. You will design and run the systems that deliver computational resources, and directly support the researchers who use them to ensure a stable and reliable research computing environment.

Your day-to-day will involve proactively monitoring our platforms, responding to requests, managing user accounts and budgets, and swiftly resolving technical issues across our wide array of GPU, API, and cloud platforms. As MATS scales, you will help build, streamline, and automate our compute infrastructure. You will analyze usage patterns to shape compute strategy, improve policies, and create the documentation and internal tooling that supports our growing research community.

You might be:

• IT Administrators or technical support specialists who excel at user-facing support and are looking to apply their skills in a mission-driven research environment.
• Technical operations or project professionals from fast-scaling or tech-focused organizations who are experts at building robust systems, managing complex logistics, and coordinating with stakeholders.
• Software engineers or system administrators who enjoy a mix of user interaction, problem-solving, and system-building, and want to contribute directly to research enablement.

If you do not fit perfectly into one of these profiles but believe your skills are a strong match for the responsibilities, we are still excited for you to apply!

Key Responsibilities

Scholar compute provisioning and support:

• Enable cutting-edge AI alignment, control, security, and governance research by providing seamless and responsive access to essential compute resources.
• Triage, troubleshoot, and resolve scholar requests related to compute access, account setup, and platform issues in a timely manner.
• Administer scholar accounts across various GPU and API platforms, managing top-ups, access controls, and usage limits.
• Assist scholars using the MATS Cluster, including help with account setup, job submissions, and basic troubleshooting.

System & financial administration:

• Own and manage our compute request infrastructure, including Airtable databases, and implement improvements to enhance efficiency and scalability.
• Administer and enforce security policies, including role-based access control (RBAC), API key management, and user offboarding procedures.
• Diligently track compute spending and usage for each scholar, providing budget alerts and supporting budget planning.
• Conduct regular audits of platform funds, security settings, and potential budget overages to ensure financial and operational stability.
• Review and triage special compute requests from mentors and external collaborators.

Process improvement and documentation:

• Build and improve systems and internal tooling to help our team scale effectively.
• Create and maintain clear technical documentation for our systems, policies, and standard operating procedures.
• Assist in researching, vetting, and onboarding new compute platforms that meet MATS's research, security, and financial criteria.

Essential Qualifications and Skills

• 2+ years of experience across a combination of the following:
  • IT administration
  • Technical support
  • Operations in a technical context
  • Program coordination in a technical context
• Ability to create and maintain clear technical documentation.
• Excellent customer service mindset, with a genuine desire to help people succeed.
• Comfort and basic familiarity with a command-line environment (e.g., Linux terminal), including navigation, file operations, and using SSH.
• Proficiency with data management tools, including databases (e.g., Airtable) and spreadsheets (e.g., Google Sheets).
• Strong technical problem-solving skills, with a proven ability to quickly learn, administer, and troubleshoot various web-based platforms.
• Proven ability to manage a high volume of concurrent tasks and requests with strong attention to detail.
• Proactive and self-driven work ethic.
• Alignment with our mission of reducing risk from the development of advanced AI.

Desirable Qualifications and Skills

We expect especially strong applicants to have deep experience in at least one of the following areas:

• Experience in a dedicated IT Admin or Technical Support role.
• Experience with financial administration, including budget tracking, expense monitoring, and reporting.
• Experience with scripting (e.g., Python, Bash) or software engineering for task automation and building internal tools.
• HPC / Cluster Administration experience:
  • Linux system administration
  • Job scheduler/Workload management (e.g., Slurm)
  • Container technologies (e.g., Docker, Apptainer)
  • Storage and networking (e.g., NFS, distributed file systems)
• Familiarity with Identity and Access Management (IAM) principles and tools (e.g., SSO, MFA, RBAC).
• Basic familiarity with cloud administration on major providers like AWS or GCP.
• Experience with helpdesk or ticketing software (e.g., Zendesk, Jira).
• Proficiency with AI tools (e.g., Claude, Cursor) for automating workflows or generating simple scripts.
• Strong competence with Airtable, Google Suite, Slack, and Notion.
• Familiarity with the basic ideas of AI alignment and security.

What We Offer

• Opportunity to help accelerate and launch the careers of talented AI alignment, control, security, and governance researchers
• Professional development and skill enhancement opportunities
• Flexibility for remote work between cohorts
• Collaborative and intellectually stimulating work environment
• Access to AI safety-focused office spaces in Berkeley
• Medical insurance, including dental, vision, and life insurance
• Traditional and Roth 401(k) options
• Meals provided Monday-Friday
• Visa sponsorship may be possible but is not guaranteed; we encourage all interested candidates to apply

Compensation

Compensation will be $130,000 to $200,000 annually, depending on experience.

Working hours and location

40 hours per week. Successful candidates can expect to spend most of their time working in-person from our main office in Berkeley, California. We are open to hybrid working arrangements for exceptional candidates. The role includes participation in a rotating weekend on-call schedule of typically 1-2 times per month, to provide support for time-sensitive scholar needs, such as key deadlines or critical platform issues. This responsibility requires general availability and responsiveness during the on-call period, though in practice the active workload has historically been light and spread out across the weekend.

How to Apply

Please fill out the form here. If you use an LLM chatbot or other AI tools in this application, please follow these norms. Applications will be reviewed on a rolling basis.

MATS is committed to fostering a diverse and inclusive work environment at the forefront of our field. We encourage applications from individuals of all backgrounds and experiences.

Join us in shaping the future of AI alignment, control, security, and governance research!