Data Scientist, AI Model Risk

Job Summary

Job Description

What is the opportunity?

RBC is a global leader in applying Artificial Intelligence (AI) in the banking sector in order to create value for our clients, with capabilities ranging from LLM-powered digital banking, boosting ensembles in fraud detection and AML, voice assistants in customer service, to algorithmic trading in capital markets. A failure to effectively prepare for and manage emerging model risk related to AI would subject RBC to financial, regulatory, and reputational risks and, as a result, RBC would not be able to provide its clients with the best quality service. Therefore, the AI validation team within RBC’s Enterprise Model Risk Management (RBC Group Risk Management) is tasked with overseeing, assessing, and managing the model risk that may arise from these AI capabilities.

The AI validation team uses machine learning, statistical, and computational strategies to assess model risk. In doing so, RBC is able to identify model weaknesses early and enhance the reliability of production models across all lines of business.

What will you do?

• Application: You will have the opportunity to work in any of the many areas we work in, across an even wider variety of business functions, such as the following:

Types of Models: Classification, regression, anomaly detection, natural language processing, computer vision, reinforcement learning, recommendation systems, dimensionality reduction, Large Language Models including generative AI.

Business Functions: Internal Audit, Cybersecurity, Fraud Management, Anti-Money Laundering, Insurance, Credit Risk, Technology Operations, Identity & Access Management, Human Resources.

• Validation: Your role is to challenge models and identify risks associated with their use – both conceptually and empirically. To that end, you will design and execute validation frameworks, exploring modelling considerations such as conceptual soundness, data processing, metric reproducibility & stability, benchmarking, robustness, uncertainty quantification, fairness, privacy, explainability, implementation controls and more. You will also have the freedom to explore ideas that interest you and build your own models and tools.

• Research & Development: You will read research papers (established work and state-of-the-art) to enhance how our team validates models and contribute to our knowledge pool. You are encouraged to apply what you’ve learned to real-world problems, develop reusable software packages, and share your insights with others.

• IT: You will collaborate with cross-functional stakeholders to establish and promote best-practices related to MLOps, tooling and IT infrastructure.

• Governance: You will work with model developers (data scientists, researchers, engineers) and business stakeholders to inventory applications of AI and machine learning at the bank, determine their materiality, and assess whether they require review.

What do you need to succeed?

Must-have

• Passionate about learning and staying up-to-date with research and technology
• Strong communication and interpersonal skills
• Progress towards a PhD or Master’s degree in Statistics, Computer Science, Applied Mathematics, Econometrics, Engineering, Quantitative Finance, or a related quantitative field
• Proficient programming skills in Python or a similar language; you should already be comfortable with writing research experiments and be willing to learn how to write clean code.
• Familiarity with popular machine learning frameworks and libraries
• Publications or GitHub portfolio is must to have.

Nice-to-have

• A risk-oriented mindset: You are curious about the “how” as well as the “why