The Head of Engineering will build and scale the engineering team and systems to execute high-stakes red-teaming engagements for frontier AI models. This role involves translating technical strategy into scalable execution, mentoring staff, and developing tools to identify vulnerabilities in AI systems.
You will operate and scale the Kubernetes GPU fleet, managing node lifecycles, storage, and batch scheduling for research experiments. Additionally, you will collaborate directly with research teams to optimize infrastructure performance, fault tolerance, and security.
You will lead the infrastructure sub-team to own the technical direction, roadmap, and operations of the GPU cluster fleet. You will work directly with researchers to ensure high-performance, fault-tolerant experiments while hiring and coaching a team of senior engineers.
You will lead a research team to develop and implement pre-training safety methods for large-scale AI models, focusing on capability control and harm reduction. Additionally, you will define the research agenda, mentor staff, and collaborate with external partners to drive real-world adoption of safety standards.
Define and lead technical programs for foundational research and infrastructure teams, taking complex projects from idea to delivery. Coordinate between researchers, engineers, and internal operations to optimize resource roadmaps and facilitate technology transfer to external stakeholders.
Build and scale the entire people function, managing the employee lifecycle from talent acquisition to retention. Act as a strategic partner to leadership on organizational design, compensation, and culture for a growing AI research institute.
Own the end-to-end planning and delivery of a high volume of concurrent events, including logistics, vendor coordination, and budget tracking. Coordinate dependencies across functions and support junior project managers to build team autonomy and operational systems.
Lead the intellectual design and curation of a diverse portfolio of AI safety events, including agenda creation and speaker selection. Manage pre-event content framing and post-event continuity to ensure long-term impact and field-building.
Lead the technical jailbreaking practice by discovering high-severity vulnerabilities in frontier AI models and dismantling defense-in-depth stacks. Translate these findings into real-world mitigations by collaborating with frontier labs and government agencies.
The Research Lead will define and own a research workstream end-to-end, articulating a research agenda to mitigate catastrophic risks from advanced AI. They will also grow and lead a team of technical staff, mentor junior members, and drive adoption of research findings through publications and partnerships.
