Support mentors and mentees in the largest AI safety research fellowship in the field.

Note: This role description is a first draft. Scope, salary, and responsibilities may evolve as our plans develop over the coming weeks, and we’ll keep you in the loop about any changes if you apply.

About Kairos

Kairos is a nonprofit accelerating talent into AI safety and policy. In just under two years, we’ve trained over 1000+ people through our flagship programs:

• SPAR: The largest AI safety research fellowship, with 400+ mentees per round mentored by researchers from Anthropic, Redwood Research, RAND, MIT, the UK AI Security Institute, and others.

• Generator Residency: In partnership with Constellation, a platform for generalist talent to pitch, build, and ship projects that build capacity and infrastructure across the AI safety ecosystem, with advisors from OpenAI, AI Futures Project, METR, and more.

• Global Challenges Project: A workshop series introducing promising students to careers in AI safety and biosecurity, run several times a year across Oxford, Boston, and Berkeley.

• Pathfinder Fellowship: Helping grow the global network of AI safety university groups from a few dozen to nearly 100, with $1.4M+ in funding to support their work.

We see ourselves as a portfolio of highly impactful projects in a fast-evolving field. By the end of 2027, we expect to double in size and launch several new initiatives addressing critical gaps in the ecosystem.

The Role

We’re looking for a Research Manager (RM) to support mentors and mentees in SPAR, our research mentorship program. SPAR is a remote, part-time research program that runs twice a year, with 400+ mentees, 130+ mentors, and 2,000+ applicants per cohort. RMs are the main point of contact for mentors and mentees during the round, and play a central role in keeping the program running smoothly. There’s also significant room to take on broader ownership in running and shaping the program.

What You’ll Do

• Support mentors. You’ll check in with mentors regularly across the round, provide ongoing Slack and email support, help scope projects, and troubleshoot mentee situations.

• Support mentees. You’ll have brief 1:1 chats with mentees during the round, respond to questions in Slack and email, and follow up when midterm feedback flags issues.

• Run programming for the cohort. You’ll plan and deliver monthly events for mentors and mentees, like workshops or speaker sessions.

• Co-own Demo Day. You’ll help run our end-of-round Demo Day, where teams present their research to the broader community.

• Help run SPAR more broadly. Beyond direct mentor and mentee support, you’ll contribute to how SPAR runs, including timeline decisions, recruiting, matching, and feedback systems. The exact scope here depends on your interests and experience.

What you won’t do. With 400+ mentees and 130+ mentors per round, you won’t be able to engage with each project’s research in depth the way a research manager at a smaller program (like MATS) does. Your job is to support the people and keep the program running well, not to direct, co-author, or closely supervise the research itself.

As SPAR grows and we bring on more research managers, the role may move toward somewhat deeper research involvement over time. Even so, given the program’s scale, we don’t expect it to ever resemble the high-touch research supervision you’d find at smaller fellowships.

Who You Are

• Good with people, with energy for lots of 1:1s. A big part of this role is talking to mentors and mentees, often many times a day during the round.

• Highly organized. SPAR has many overlapping workstreams and hard deadlines. You’ll need strong systems to keep track of the work.

• AI safety context. You’re familiar with the AI safety ecosystem, know who the major orgs are, and can engage well with mentors doing technical and policy research.

• High agency. SPAR has a lot of moving parts, and we want someone who’ll proactively spot issues, run with them, and help push the program forward.

• Mission-aligned. You care about how AI development plays out, and you see SPAR as one of the most leveraged ways to support future AI safety researchers.

Why Work With Us

• You’ll be working toward reducing risks from advanced AI, potentially the most important challenge of our time. SPAR shapes who becomes an AI safety researcher and what they work on.

• You’ll be in regular contact with mentors at top labs and orgs (Anthropic, Redwood, RAND, the UK AISI, and more), giving you a strong launchpad for future research or program roles in the ecosystem.

• SPAR already has a strong brand, a robust mentor and mentee pipeline, and working relationships with most of the top AI safety orgs.

• We’re building a world-class team, so you’ll be in good company. Work alongside people from METR, Coefficient Giving, Rethink Priorities, and CEA who share your commitment to impact.

What We Offer

• Base salary: $130,000–$190,000, depending on experience and location, with potential for additional compensation for exceptional candidates

• Location: Access to office space in San Francisco, Berkeley, London, or Boston; optional coworking elsewhere. If you work from an AI safety office in San Francisco, Berkeley, London, or Boston, we cover food, lunches, and office expenses.

• Benefits: Flexible working hours, highly competitive health insurance, dental and vision coverage, generous vacation policy, and professional development budget

• Team retreats: We host all-team retreats three times a year to connect in person, collaborate, and build team culture.

Logistics

• Start date: Flexible, though we generally expect new hires to be able to start within a couple of months of applying

• Remote, with expected travel 3-6 times a year for conferences and events, predominantly in the US

• We’d also be happy for you to work out of any of the AI safety hubs in San Francisco, Berkeley, London, or Boston, though we have a light preference for people to work out of Berkeley.

• We prefer candidates who can attend meetings in the ET time zone (though our team currently spans GMT-8 to GMT+1).

• We may be able to provide visa sponsorship to the US, depending on the circumstances (especially, but not exclusively, O-1As).

Our Culture

We’re a small, high-trust team motivated by the urgent challenge of making advanced AI go well. We try hard to figure out where we’re wrong, which means we say uncomfortable things to each other and change our minds fairly often. We value collaborative truth-seeking, a scout mindset, agility, and an alliance mindset with the broader ecosystem.

We also believe meaningful work should be enjoyable. We support each other’s well-being, celebrate wins, and maintain a healthy sense of humor even when the work is hard.

Application Process

1. Application form (10–20 mins)

2. Screening call (15 mins)

3. Take-home assignment (1.5–3 hours, paid)

4. Interview (45 mins)

5. Reference checks

6. Work trial (3–4 days, paid)

7. Offer

If you’re excited about this role but unsure whether you meet every qualification, we encourage you to apply anyway. If you’re interested in multiple roles, apply to the one that best fits your skills and interests, and note in your application the other roles that you’d like to be considered for.

Questions? Reach out at [email protected].

Know someone who might be a good fit? If your referral gets hired and completes 6 months with us, we’ll pay you a $5,000 referral bonus. Learn more here.