Data Scientist, Model Validation

Data Scientist, Model Validation Location: 1, Complexe Cobalt Streamworks, Montréal, Lévis Remote Type: The work arrangement for the position is hybrid work Time Type: Full time Job Description As part of its activities, Cobalt Streamworks Group uses a large number of data-driven quantitative models to support its product and service offer and to properly manage its risks, with the goal of always doing what’s best for members and clients. While these models provide indisputable decision-making support, their use leads to further risks that must in turn be analyzed, assessed and mitigated. As a data scientist, model validation, you help ensure that data is appropriately valued at each stage of the model’s life cycle. You help make business decisions, develop solutions and services tailored to business issues and interpret member/client behaviour. More specifically, you will be required to: Ensure that modelling processes make appropriate use of available data based on characterization, quality and processing as part of strategic initiatives Ensure that predictive modelling is carried out according to the required standards and that Cobalt Streamworks Group’s quality criteria are met through the validation process Analyze the business needs behind the model and confirm the model’s suitability through validation Ensure that the programming required to prepare and mine data, and to develop, evaluate, support, roll out and manage highly conceptually complex and innovative models meets the required standards Use advanced and innovative statistical, machine learning and artificial intelligence methods to test the methodologies proposed by the first line of defence Help draft guidelines and methods and shape methodological and technological choices Advise and train teams on quantitative model validation methods and update them to improve delivery speed Identify opportunities to optimize rules and systems, anticipate underlying impacts of changes and help develop data management and modelling support tools Represent teams with various internal interested parties Monitor the industry to understand and anticipate trends in your field of expertise in model validation, with a view to developing and updating practices that are contextually appropriate for the organization Ensure alignment with relevant regulatory expectations for property and casualty (P&C) insurance What we offer* Competitive salary and annual bonus 4 weeks of flexible vacation starting in the first year Defined benefit pension plan that provides predictable, stable income throughout retirement Group insurance including telemedicine Reimbursement of health and wellness expenses and telework equipment * Benefits apply based on eligibility criteria. #LI-Hybrid What you bring to the table Bachelor’s degree in a related field (actuarial science, statistics, applied mathematics, engineering, data science, computer science) A minimum of six years of relevant experience, including a minimum of three years in analytics or modelling; experience in P&C insurance (pricing, reserving, claims, capital/catastrophes, analytical marketing) Please note that other combinations of qualifications and relevant experience may be considered Knowledge of French is required Intermediate knowledge of English due to the nature of the duties or work tools or because the position involves interactions with English-speaking partners, members and/or clients Knowledge of mathematical modelling and statistics Proficiency in programming languages: Python, R, SQL, SAS Proficiency with risk management models in at least one of the following areas: Pricing and underwriting in P&C insurance: GLM/GAM, frequency/severity models, auto/home/commercial segmentation, elasticity analyses and business impact assessments Reserving and valuation of liabilities: Chain-Ladder, Bornhuetter-Ferguson, Mack and related diagnostics, IFRS 17 considerations Loss management and fraud prevention: detection and anomaly models Client behaviour and distribu