At a Glance
- Tasks: Join our team to transform AI control research into practical safety tools.
- Company: Innovative AI safety company focused on reducing risks from AI.
- Benefits: Competitive salary, unlimited vacation, flexible hours, and professional development budget.
- Other info: Collaborative environment with opportunities for rapid growth and responsibility.
- Why this job: Make a real impact on AI safety while working with cutting-edge technology.
- Qualifications: 2+ years in empirical research with AI systems and strong Python skills.
The predicted salary is between 60000 - 80000 £ per year.
Application deadline: We are conducting interviews actively and aim to fill this role as soon as we find someone suitable.
THE OPPORTUNITY
Join our new AGI safety product team and help transform AI control research into practical tools that directly reduce risks from AI. As a Research Scientist (Control), you’ll work closely with Marius (CEO & currently leads the monitoring efforts), other control researchers and product engineers.
We are currently building Watcher, a monitoring tool for coding agents. Our monitoring research agenda attempts to translate compute into safety at scale. You will join a small team and will have significant ability to shape the team & tech, and have the ability to earn responsibility quickly.
You will like this opportunity if you're passionate about using empirical research to make AI systems safer in practice. You enjoy the challenge of translating theoretical AI risks into concrete detection mechanisms. You thrive on rapid iteration and learning from data. You want your research to directly impact real-world AI safety.
KEY RESPONSIBILITIES
TLDR: you will design & implement control protocols and test them on real-world production systems at scale.
- Research & Development
- Systematically collect and catalog coding agent failure modes from real-world instances, our internal deployments, public examples, research literature, and theoretical predictions
- Design and conduct experiments to test monitor effectiveness across different failure modes and agent behaviors
- Build and maintain evaluation frameworks to measure progress on monitoring capabilities
- Build and maintain high-quality datasets to train and test monitors on
- Iterate on monitoring approaches based on empirical results, balancing detection accuracy with computational efficiency
- Stay current with research on AI safety, agent failures, and detection methodologies
- Stay current with research into coding security and safety vulnerabilities
- Monitor Design & Optimization
- Develop & maintain a comprehensive library of monitoring prompts tailored to specific failure modes (e.g., security vulnerabilities, goal misalignment, deceptive behaviors)
- Experiment with different reasoning strategies and output formats to improve monitor reliability
- Design and test hierarchical monitoring architectures and ensemble approaches
- Optimize log pre-processing pipelines to extract relevant signals while minimizing latency and computational costs
- Implement and evaluate different scaffolding approaches for monitors, including chain-of-thought reasoning, structured outputs, and multi-step verification
- Fine-tuning & Red-teaming
- Fine-tune open-source models to create efficient monitors for high-volume production environments
- Design and build agentic monitoring systems that autonomously investigate logs to identify both known and novel failure modes
- Build automated red-teaming pipelines that attack monitors at scale
- Design iterative adversarial games where a red-team and blue team continuously attack and defend respectively
JOB REQUIREMENTS
- 2+ years of experience conducting empirical research with large language models or AI systems
- Strong experience with AI coding agents
- Experience with LLM-as-a-judge setups or AI monitoring more broadly
- Experience designing and running experiments, analyzing results, and iterating based on empirical findings
- Strong Python programming skills
- Demonstrated ability to work independently on open-ended research problems
Bonus:
- Familiarity with AI safety concepts, particularly agent-related risks
- Familiarity with computer security, e.g. security testing and secure system design
- Experience fine-tuning language models or working with smaller open-source models
- Previous work building developer tools or monitoring systems
- Publications or contributions to AI safety or ML research
- Experience with production log systems or production log analysis
We want to emphasize that people who feel they don't fulfill all of these characteristics but think they would be a good fit for the position nonetheless are strongly encouraged to apply. We believe that excellent candidates can come from a variety of backgrounds and are excited to give you opportunities to shine.
WHAT YOU'LL ACCOMPLISH IN YOUR FIRST YEAR
- Build a comprehensive failure mode database: Systematically collect and categorize 100+ distinct AI agent failure modes across safety and security dimensions, creating the foundation for our monitoring library.
- Develop and validate monitoring approaches: Create and empirically test monitoring prompts and strategies for key failure categories, establishing clear metrics for monitor performance and building evaluation frameworks to track progress.
- Optimize the monitoring pipeline: Improve log preprocessing and monitor scaffolding to achieve measurable improvements in detection accuracy, false positive rates, and computational efficiency.
- Advance monitoring capabilities: Work on advanced approaches such as fine-tuned monitors or agentic investigation systems.
REPRESENTATIVE PROJECTS
- Hierarchical monitoring for coding agent security: Design a multi-layer monitoring system for detecting security vulnerabilities introduced by coding agents.
- Design a backtesting strategy for Watcher: Validate the system on synthetic test cases and real agent outputs, iterating to optimize the tradeoff between detection rates and false positives while maintaining low latency for most monitoring decisions.
- Fine-tune an open-source model to be a great monitor: Take an open-source model and fine-tune it on our large dataset of coding agent failures with high-quality ground truth labels.
BENEFITS
This role offers market competitive salary, equity, and competitive benefits.
- Salary: 100k - 200 GBP (~150k - 270k USD)
- Flexible work hours and schedule
- Unlimited vacation
- Unlimited sick leave
- Up to 6 months of paid parental leave
- Comprehensive health, dental and vision insurance
- Retirement savings with competitive employer matching
- Lunch, dinner, and snacks are provided for all employees on workdays
- Paid work trips, including staff retreats, business trips, and relevant conferences
- A yearly $1,000 (USD) professional development budget
- Relocation support and visa fees (if applicable)
LOGISTICS
- Time Allocation: Full-time
- Location: This is an in-person role working out of our London or San Francisco office. We offer flexible working hours and wfh arrangements.
- Visa sponsorship: We sponsor visas in both the UK and US.
ABOUT THE TEAM
The Product team is a new team. Especially early on, you will work closely with Marius Hobbhahn (CEO & currently leads the monitoring team), Victor Gillioz (Research Scientist), Monika Jotautaitė (Research Scientist), and our product engineers.
ABOUT APOLLO RESEARCH
The rapid rise in AI capabilities offer tremendous opportunities, but also present significant risks. At Apollo Research, we’re primarily concerned with risks from Loss of Control, i.e. risks coming from the model itself rather than humans misusing the AI.
We work on the detection of scheming, the science of scheming, and scheming mitigations. We closely work with multiple frontier AI companies, e.g. to test their models before deployment or collaborate on scheming mitigations.
At Apollo, we aim for a culture that emphasizes truth-seeking, being goal-oriented, giving and receiving constructive feedback, and being friendly and helpful.
EQUALITY STATEMENT
Apollo Research is an Equal Opportunity Employer. We value diversity and are committed to providing equal opportunities to all, regardless of age, disability, gender reassignment, marriage and civil partnership, pregnancy and maternity, race, religion or belief, sex, or sexual orientation.
HOW TO APPLY
Please complete the application form with your CV. The provision of a cover letter is neither required nor encouraged. Please also feel free to share links to relevant work samples.
About the interview process: Our multi-stage process includes a screening interview, a take-home test, technical interviews, and a final interview with Marius (CEO).
Your Privacy and Fairness in Our Recruitment Process: We are committed to protecting your data, ensuring fairness, and adhering to workplace fairness principles in our recruitment process.
Research Scientist (Control) employer: Apollo Research
Apollo Research is an exceptional employer, offering a unique opportunity to shape the communications function of a leading AI safety organisation from the ground up. With a commitment to truth-seeking and a collaborative work culture, employees benefit from flexible hours, unlimited vacation, and a generous professional development budget, all while making a meaningful impact in the field of AI safety. Located in San Francisco, the role provides strategic autonomy and the chance to engage with top-tier US media, ensuring that your expertise reaches vital audiences.
StudySmarter Expert Advice🤫
We think this is how you could land Research Scientist (Control)
✨Get Involved in Control Systems Communities
Dive into control systems forums and relevant LinkedIn groups to showcase your expertise. Engage in discussions, answer questions, and share your knowledge—this can help us get noticed by potential employers looking for someone just like you!
✨Attend Industry Conferences and Workshops
Join conferences or workshops specifically tailored to control systems engineering, like the IEEE conference. These events are goldmines for networking with industry leaders and might just lead us to that perfect role at Apollo Research.
✨Get Practical Experience Through Projects
Start or contribute to projects that highlight your control systems skills. Whether it’s a simulated environment or real-world application, having tangible results will give us a huge edge when chatting with Apollo Research.
✨Choose Job Boards Wisely
Focus on niche job boards that specialise in engineering roles. Channels like Engineers.co or similar can help ensure that we’re not only applying to random jobs but targeting positions that align perfectly with our control systems engineering skills!
We think you need these skills to ace Research Scientist (Control)
Some tips for your application 🫡
Highlight Your Technical Skills:When applying for the Research Scientist (Control) role, make sure to showcase your technical expertise in control systems. Be specific about the software and programming languages you’ve worked with—mention anything from MATLAB to PLC programming. This will help us see your proficiency in the tools that are essential for the job.
Quantify Your Achievements:In the engineering world, numbers speak volumes! Use your CV to quantify your achievements—whether it’s a percentage improvement in system efficiency or successful project completions under budget. This will not only make your application stand out but also showcase your problem-solving capabilities to us.
Tailor Your Cover Letter:Don’t hesitate to get personal in your cover letter! Talk about why you’re passionate about control systems engineering and how your previous experiences have prepared you for this role at Apollo Research. We want to see your genuine interest and motivation to contribute to our projects.
Professional Documentation:For a full-time role, ensure that your application documents are professional and well-structured. Include a clear CV format and cover letter that highlights key projects or systems you've worked on. Don’t forget to mention any relevant certifications or training that would make you a valuable addition to us at Apollo Research.
How to prepare for a job interview at Apollo Research
✨Brush Up on Control Systems Concepts
Make sure we're up to speed with control systems theory—think stability analysis, PID tuning, and state-space representation. Revising these fundamentals will help us answer technical questions more confidently and demonstrate our expertise during the interview.
✨Familiarise Yourself with Simulation Tools
Get hands-on with popular simulation tools like MATLAB/Simulink or LabVIEW. A lot of interviews might include technical assessments or practical tasks to gauge our understanding of system modelling and control design using these tools, so being proficient could really set us apart.
✨Be Ready for Problem-Solving Scenarios
We should expect scenario-based questions where we'll have to solve real-world control system challenges. Practising how to approach problems logically and articulately will help us show our analytical thinking to the interviewer at Apollo Research.
✨Show Your Passion for Engineering
As this role is full-time, it's crucial we convey our genuine enthusiasm for control systems engineering. Sharing personal projects or experiences where we tackled interesting engineering tasks can illustrate our commitment and eagerness to contribute to the team at Apollo Research.