Director, Model Risk — AI & Fraud

 

 

• →
  Establish and own the independent validation framework for fraud models and AI-enabled decision systems across OKX products, covering conceptual soundness, data quality, implementation, and outcomes analysis, in line with SR 26-2 and equivalent model risk frameworks.
• →
  Assess model conceptual soundness, intended use, assumptions, and limitations.
• →
  Personally execute end-to-end model validations as the technical expert, including independently reproducing results through hands-on data analysis and code review.
• →
  Assess AI/ML and generative AI applications, including LLM-based and agentic systems, for robustness, reliability, explainability, and appropriate human oversight, applying strong applied AI evaluation expertise.
• →
  Perform ongoing monitoring review and periodic revalidation to ensure sustained model performance through changing fraud patterns and business conditions.
• →
  Provide effective challenge to model owners; communicate conclusions, limitations, and risk ratings clearly to senior management and governance forums, and track remediation through closure.
• →
  Build and continuously mature the fraud model validation capability, standards, and workflows within MRM.

 

• Master’s or Ph.D. in a quantitative field such as Statistics, Mathematics, Computer Science, or a related discipline, or equivalent experience.
• 10+ years of quantitative, fraud risk, or model risk experience, with a recognized track record as a hands-on technical expert who personally delivers independent validations of models and decision systems.
• Excellent communication and judgment, including the ability to explain quantitative findings to senior audiences and constructively challenge model owners.
• Practical experience with statistical and machine learning methods; proficiency in Python and SQL; working knowledge of SR 11-7 or equivalent model risk and AI governance frameworks.
• Experience with AI/ML governance and controls, including explainability, fairness, robustness, ongoing monitoring, or generative and agentic AI applications in fraud operations.

 

• Experience establishing validation frameworks, methodologies, or capabilities in settings where they did not previously exist.
• Strong applied AI capability: hands-on experience evaluating or building ML/LLM-based systems in production, including prompt engineering, evaluation harnesses, fine-tuning assessment, or agentic AI workflows; able to run red-team style testing on AI applications.

 

OKX is committed to equal employment opportunities regardless of race, color, genetic information, creed, religion, sex, sexual orientation, gender identity, lawful alien status, national origin, age, marital status, and non-job related physical or mental disability, or protected veteran status. Pursuant to the San Francisco Fair Chance Ordinance, we will consider for employment qualified applicants with arrest and conviction records.