Armilla AI

Applied Scientist, AI Risk

Toronto, Canada

Sponsorship not specifiedDetected 23 days ago

PythonMachine LearningDeep LearningTensorFlowPyTorchscikit-learnPandasNumPyLLMsStatisticsResearchProblem SolvingActuarial ScienceUnderwriting

About the role

We're seeking an exceptional Applied Scientist who bridges the worlds of deep AI research and practical, production-grade applications.
As our Applied Scientist, AI Risk, you'll be instrumental in both advancing our understanding of AI systems and translating that knowledge into robust risk assessment and evaluation frameworks.
You'll be shaping how the insurance industry evaluates and prices AI risk.

Unparalleled opportunities to develop expertise at the intersection of AI research, risk management, and insurance alongside deeply experienced AI and industry experts.
Design and develop AI systems-including specialized models, agents, and automated evaluation pipelines-that assess the safety, reliability, and risk profiles of other AI systems.
Build production-grade tooling and platforms for automated AI risk assessment, model testing, and continuous monitoring.
Develop novel evaluation methodologies and metrics that capture AI-specific risks such as adversarial vulnerabilities, distribution shift, hallucinations, and behavioral misalignment.
Collaborate closely with our underwriting and actuarial teams to translate technical findings into actionable risk insights and pricing signals.
Meta-AI Challenge: Tackle the fascinating problem of building AI systems that understand and evaluate other AI systems.
Professional Growth: Unparalleled opportunities to develop expertise at the intersection of AI research, risk management, and insurance alongside deeply experienced AI and industry experts.

Strong track record of applied research-you've published, contributed to open source, or shipped ML products that had real-world impact beyond academic settings.
Hands-on experience with model evaluation, testing, and validation-you think critically about where models fail, not just where they succeed.
Solid software engineering skills with expertise in Python and experience with ML frameworks (PyTorch, TensorFlow, JAX) and scientific computing libraries (NumPy, Pandas, Scikit-learn).
Experience with LLMs and generative AI, including familiarity with their unique risks, evaluation challenges, and safety considerations.
Ability to work both independently on deep technical problems and collaboratively in a fast-paced startup environment.

This listing is sourced directly from Armilla AI's careers page and normalized into a canonical job model.