Mercor seeks a Radiological Safety Expert to assess AI models for misuse potential. This writing-focused role involves prompt creation, response evaluation, and ensuring safety in technical queries. As a member of Mercor's safety panel, your expertise is crucial in differentiating benign from adversarial requests.

You will write detailed prompts, judge AI responses based on policy standards, and provide comprehensive reference answers. Candidates should hold a relevant license and possess a strong background in radiation safety. Key Responsibilities:

- Develop single-turn prompts across three classification levels
- Critically evaluate AI responses against established policy standards
- Document reference answers with explicit technical reasoning




- Identify dual-use scenarios relevant to radiological safety
- Collaborate with a team of radiological safety experts Requirements:
- Relevant license holder with radiological safety experience
- Background in research, compliance, or program management
- Strong technical writing and communication skills
- Ability to analyze potentially harmful inquiries
- Experience in red-teaming scenarios or similar assessments Bring your expertise in radiological safety and technical writing to help guide AI models effectively.

📌 Radiological Safety Expert for AI Models (Ottawa)
🏢 Obsidian
📍 Ottawa