Applied Scientist, AI Risk

Responsibilities

• Conduct deep technical research into AI model behavior, failure modes, and edge cases
• Design and develop AI systems, including specialized models and automated evaluation pipelines, to assess the safety and reliability of other AI systems
• Build production-grade tooling for automated AI risk assessment and continuous monitoring
• Evaluate diverse AI systems from traditional ML to LLMs and multimodal systems
• Develop novel evaluation methodologies for adversarial vulnerabilities, distribution shift, and hallucinations
• Collaborate with underwriting and actuarial teams to translate technical findings into risk insights
• Write clean, maintainable, and well-tested code for research prototypes and production systems

Requirements

• Advanced degree (Master's or PhD) in Computer Science, Machine Learning, Statistics, or related field
• Demonstrated research experience in AI/ML
• Strong track record of applied research, open source contributions, or shipping ML products
• Deep expertise in machine learning fundamentals, classical ML, and deep learning
• Experience building AI/ML systems from conception to deployment
• Hands-on experience with model evaluation, testing, and validation
• Proficiency in Python and ML frameworks such as PyTorch, TensorFlow, or JAX
• Experience with LLMs and generative AI safety considerations

Preferred Qualifications

• Background in AI safety, robustness, interpretability, or adversarial ML
• Experience with scientific computing libraries like NumPy, Pandas, and Scikit-learn

About the Company

Armilla AI is a cutting-edge startup at the intersection of artificial intelligence and insurance. Based in Toronto, we build innovative solutions to manage, underwrite, and insure the risks associated with evolving AI systems.