Agentic AI Safety Research Scientist

Amii (Alberta Machine Intelligence Institute); Edmonton; On-site; CAD 120,000 - 180,000; Full time

Benefits offered by this job

Professional development

Open research collaboration

Conference funding

Job summary

Amii, the Alberta Machine Intelligence Institute, Edmonton, Canada, is seeking an Applied Research Scientist to advance agentic AI safety. You will lead research on autonomous AI risks, develop evaluation methods and practical tools, and publish findings.

You will report to the Director, AI Trust and Safety, collaborating with researchers and engineers to strengthen AI safety in real-world applications.

Qualifications

• PhD in Machine Learning, AI, CS or related field.
• 2+ years post-PhD research in applied/industry ML.
• Experience with AI safety research and evaluation methodologies.
• Strong knowledge of modern AI/ML, including agentic AI and tool use.
• Experience leading research projects from question to publication/deployment.
• Proficiency in Python and ML frameworks; familiarity with experimentation tooling.

Responsibilities

• Develops a clear research vision aligning with Amii's Trust and Safety program.
• Proposes and leads agentic AI safety research projects and funding proposals.
• Develops experiments, evaluation pipelines, and tools with engineering collaboration.
• Translates findings into practical evaluation tools, frameworks, and applications.
• Authors peer‑reviewed papers, whitepapers, and scientific outputs.
• Represents Amii at conferences and industry events; mentors junior staff.
• Collaborates across teams to ensure project success and timely delivery.
• Supports collaborations with national/international research institutions.

Skills

Agentic AI safety

Research leadership

Python

ML frameworks

Academic publishing

Education

PhD in ML/AI/CS

Masters with industry experience

Tools

PyTorch

TensorFlow

JAX

Job description

Amii, the Alberta Machine Intelligence Institute, Edmonton, Canada, is seeking an Applied Research Scientist to advance agentic AI safety. You will lead research on autonomous AI risks, develop evaluation methods and practical tools, and publish findings.

You will report to the Director, AI Trust and Safety, collaborating with researchers and engineers to strengthen AI safety in real-world applications.