at Bank of America in Wilmington, Delaware, United States
DUTIES: Validate and peer-review complex quantitative models (including AI/ML) to ensure they follow sound modeling practice and are in incompliance with Model Governance Policy, Guidelines, and regulatory guidance and requirements. Utilize advanced mathematical and statistical approaches (e.g., generalized addictive models, time-series models, classic and Bayesian logistic regression models, optimization models), and AI/ML techniques (e.g., random forest, gradient boosting tree, neural network and other AI/ML techniques) to replicate, validate and test models. Conduct quantitative analytics and complex modeling projects, such as developing new models, analytic processes, and system approaches, to influence strategic direction and develop tactical solutions or plans. Assess model conceptual soundness, model specification and performance, underlying data, assumptions, limitations, variable selection, development process, model implementation, and documentation; Create documentation for all validation activities using LaTeX and working with Technology staff to design systems to run developed models. Provide model risk life-cycle management to ensure its fitness to business strategy and technical platform, through model identification, model risk rating, ongoing monitoring, annual model review, required action items review, model developers review validation, model health monitoring, etc. Work cross-functionally to enforce processes and to integrate more effective model validation processes; communicate issues identified through validations to businesses and governance and control functions, escalate model use breaches and remediation plans to relevant governance committees, and address any audit or regulatory concerns. Use advanced modelling and data management software, including Python, R, SAS, Xeno, H2O, and SQL to perform model replication, performance testing, bias testing, sensitivity testing and stress testing. Perform predictive modeling or independent model validation with advanced coding language and packages including Python, R, SAS, Xeno, H2O, and SQL in consumer and small business banking industry. Utilize advanced mathematical and statistical approaches (including generalized additive models, time-series models, classic and Bayesian logistic regression models, optimization models), and AI/ML techniques (including random forest, gradient boosting tree, neural network) to develop, replicate, validate and test models. Review and validate complex quantitative models for its conceptual soundness, quantitative rigor, data quality, model performance, stress testing and bias testing, to ensure they follow good modeling practices and are in compliance with regulatory guidance and requirements. Write structured, comprehensive model validation reports and other technical papers in LaTeX during model life-cycle risk management, with respect to financial regulations and model specific requirements, and present analysis and results to financial regulatory agencies. Manage large-scale datasets, and analyzing them utilizing data mining and data visualization techniques. Utilize Git/ Bitbucket to manage the version control of various validation testing codes and LaTeX codes.