Research Engineer-AI Alignment

The successful candidate will design and build environments and evaluations that help identify problematic behaviors in advanced AI systems. You will work closely with LLMs and AI agents and take ownership of projects from ideation through implementation, measurement, and refinement.

Key responsibilities

• Build environments that identify problematic behaviors in advanced AI systems.
• Design and iterate AI evaluation methodologies.
• Own projects end-to-end, from ideation to implementation and refinement.
• Work extensively with LLMs and AI agents.
• Analyze model behavior and develop effective evaluation approaches.
• Quickly learn new technical domains and apply them to research environments.
• Develop and maintain software primarily using Python.

Required qualifications

• STEM degree preferred; equivalent technical experience and strong self-taught candidates will also be considered.
• 1+ years of software engineering experience
• Strong generalist software engineering skills
• Python
• End-to-end project ownership
• Ability to quickly learn new technical domains
• Strong conceptual and problem-solving skills
• Genuine interest in AI alignment and AI safety

Benefits & perks

$200,000–$400,000 base salary Competitive equity Hybrid work arrangement in San Francisco Visa support may be available, including H-1B transfers, new H-1B, and TN Paid technical assessment options