About the organization

MATS finds and trains talented individuals for what we see as the world's most urgent and talent-constrained problem: reducing risks from unaligned artificial intelligence. We believe ambitious researchers from a variety of backgrounds have the potential to meaningfully contribute to alignment, control, security, and governance research. Through our research fellowship, we provide mentorship, training, professional network, and financial support to accelerate their careers and increase their impact.

Since 2021, we have trained over 630 researchers. 75% of pre-2026 fellows continue to work in AI alignment. 10% have co-founded organizations. Our fellows have produced 215+ research papers with 17,000+ citations. And we are just getting started: in the coming year we will expand our program to support even more researchers.

About the role

MATS directs tens of millions of dollars toward AI safety & security research each year, and hundreds of people pass through our programs — but our understanding of our own impact is still more qualitative than it should be. As the Impact Analyst, you will own the system that fixes that: a reliable pipeline of data on fellow career and research outcomes, and the analyses built on top of it that help us answer questions like "What is the impact of the marginal fellow/mentor?", "Which of our past research bets are strong and which are weak?", and "How fast should MATS scale and in which directions?"

In the short term, you will stand up and harden our outcomes-data pipeline: deciding what we collect and on what schedule, cleaning and verifying automatically collected data (supplemented by asking fellows, mentors, and research managers), and consolidating historical data. In the long term, you'll run the analyses that feed our biggest decisions — mentor selection, fellow admissions, and program scaling — and present findings to MATS leadership. Your work will directly shape decisions worth millions of dollars per year. We expect heavy, thoughtful use of AI tools throughout, and you'll collaborate closely with our Program Systems Associate on the backend systems where the data lives.

Who we're looking for

We welcome applications from analytical generalists who combine data skills with strong judgment. You might be:

• A data analyst or data scientist who has built and maintained real-world data pipelines and cares about decision quality, not just dashboards
• An analytical project manager or operations/strategy analyst who has owned a recurring, multi-stakeholder measurement process end-to-end
• A researcher or grantmaking/impact-evaluation analyst with experience turning messy real-world outcomes into defensible conclusions

If you don't match one of these profiles but can otherwise demonstrate that you'll excel in this role, we still encourage you to apply!

Key responsibilities

• Own the MATS outcomes-data pipeline: define what we collect about fellow career and research outcomes, set the collection schedule, and maintain the process so it runs like clockwork
• Design cleaning and verification workflows for automatically collected and self-reported data, including review by random samples of informed evaluators
• Find, clean, and consolidate historical data across our fellow-selection and mentor-selection systems, in collaboration with our Program Systems Associate
• Analyze outcomes to answer decision-relevant questions — headline among them the marginal value of fellows and mentors, and which factors predict the best outcomes
• Apply careful causal reasoning: our standard is counterfactual, marginal impact — not raw outcome counts
• Communicate findings to the teams that act on them: executive leadership (scaling and direction), mentor selection, and admissions
• Coordinate contributors to the pipeline (team members, contractors, interns) — you own the process without doing every piece yourself
• Report to the Program Director (Juan Gil).

Essential qualifications and skills

• Strong data analysis ability, including the judgment to avoid classic pitfalls (correlation vs. causation, selection effects, absolute vs. counterfactual impact)
• Skill at using LLMs and AI tools to accelerate data collection, cleaning, and analysis
• Strong project management ability: you can own a recurring process and keep it reliable
• Sufficient hands-on technical ability (e.g. scripting, SQL, or Airtable-level tooling) to pair with our systems team
• Clear written and verbal communication, including presenting analyses to leadership
• Familiarity with the AI safety & security ecosystem, and strong 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:

• Building automated data-collection or enrichment pipelines
• Impact evaluation, grantmaking analysis, or talent-program measurement
• Statistics or causal inference applied to small, messy, real-world datasets
• The AI safety ecosystem: its organizations, career paths, and research landscape

About the Compensation

• Compensation range for this role is $140,000 to $185,000 per year for Bay Area-based hires, depending on qualifications and experience. The essential skills listed above contain the basic requirements, and align with the lower end of this range.
• MATS provides in-office, catered lunches and dinners to employees 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 (but not for fully remote work).
• Collaborative and intellectually stimulating work environment.
• Medical, dental, vision, and life insurance.
• Both Roth 401(k) and Traditional 401(k) for US-based employees.
  
  

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.

Note while applying

If you use an LLM chatbot or other AI tools in this application, please follow the norms here. Applications will be reviewed on a rolling basis. MATS uses LLMs to help draft job descriptions, identify candidates to source, and transcribe interviews.

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 safety & security research!