Research Engineering Manager, Responsible Frontier AI Research
London, UK • Full-time
Posted 1mo ago
Job Location
London, UK
Tech Stack
Remote Work Policy
On-site
Employment Type
Full-time
Categories
AI Research Engineer
About the job
As a Research Engineering Manager, you will lead a team of Research and Software Engineers focused on building the engineering foundations for DeepMind's Responsibility research portfolio. This involves creating automated platforms for discovering, measuring, and mitigating jailbreaks, as well as developing red-teaming tools for subject-matter experts. The role also includes building defenses for AI agents operating on the open web and creating evaluations and tools to mitigate severe harms like CBRN risks and harmful manipulation, all within the framework of Frontier Safety. You will need to be a highly adaptable leader, capable of quickly understanding new research areas and directing engineering efforts effectively, balancing research experiments with scalable infrastructure. This position also requires strong technical leadership, including hiring, developing, and retaining an exceptional team, while partnering with various internal teams to address key responsible AI issues.
Responsibilities
- Set and communicate priorities for multiple workstreams aligned with the broader responsibility organization's goals, adjusting engineering capacity between research projects as priorities change.
- Set clear expectations for individuals, including tech leads, providing regular feedback, coaching, and fostering a high-performing team through hiring and development.
- Develop long-term technical goals and roadmaps for systems like evaluation pipelines, red-teaming tools, autoraters, and mitigation training pipelines, evolving them as frontier models advance.
- Oversee system designs and review code to address complex problems, establishing best practices for code quality, scalability, reliability, and secure data handling.
- Collaborate with Research Scientists, Subject-Matter Experts, Product, Program, Policy, and Trust and Safety teams to translate research into actionable evaluations and mitigations for model launches, presenting trade-offs and results to senior leadership.
Requirements
- Experience leading teams working on AI safety, security, or responsible AI research, including areas like adversarial robustness, red teaming, safety evaluations, or mitigation strategies for frontier models.
- Experience with automated red teaming or jailbreak discovery.
- Experience with LLM-based autoraters and evaluation design.
- Experience with safety classifiers or probes.
- Experience with prompt injection and agent security.
- Experience with evaluation of severe-harm risks such as CBRN or Harmful Manipulation.
- Master’s degree or PhD in Engineering, Computer Science, or a related technical field with a focus on adversarial machine learning or AI/ML security (preferred).
- Track record of research or engineering achievements, such as publications, open-source projects, or public benchmarks (preferred).
