Data Scientist - Fraud Modeling

• Partner closely and manage interaction with Strategy partners and Model Risk Management validators across the model lifecycle including validation, ongoing performance evaluation, and annual reviews
• Responsible for model development, validation and testing of fraud detection models, including statistical analysis, data validation and model performance evaluation using tools like Python, H2O
• Responsible for documenting data requirements, data collection / processing / cleaning, and exploratory data analysis, which may include utilizing statistical models / algorithms and data visualization technique
• Ensure compliance with regulatory requirements and industry best practices in fraud detection and model validation
• Identifies and compiles large complex data sets using a variety of tools (e.g. Hive, SQL) to help predict, improve, and measure the success of key business to business outcomes
• Appropriately assess risk when business decisions are made, demonstrating particular consideration for the firm's reputation and safeguarding Citigroup, its clients and assets, by driving compliance with applicable laws, rules and regulations, adhering to Policy, applying sound ethical judgment regarding personal behavior, conduct and business practices, and escalating, managing and reporting control issues with transparency

• 8+ years of hands-on analytical experience
• 5+ years’ experience in statistical analysis with working knowledge of at least one of the following statistical software packages: Python (Must), Hive, Spark, SAS
• Strong quantitative, analytical, and problem-solving skills
• knowledge of probability theory, statistics, mathematical finance, econometrics, numerical methods
• Strong proficiency in statistical analysis and model validation techniques. A solid grasp of traditional and advanced modeling techniques and algorithms, such as Logistic Regression, Gradient Boosting, Random Forests, etc.
• Expertise in model evaluation techniques such as ROC curves, precision-recall, KS, cross-validation, feature importance, SHAP values, etc.
• Familiarity with regulatory requirements and guidelines related to risk model validation
• Strong communication and interpersonal skills
• experience with Stakeholder management
• Strong project management and organizational skills
• ability to multi-task and meet deadlines
• Ability to work independently. Risk and control mindset: ability to ask incisive questions, assess materiality and escalate issues
• Ability to handle large volumes of transactional data. Familiarity with data extraction tools like Hive will be an advantage
• Successful candidate will have a demonstrable analytical problem solving, and the ability to deliver projects in a fast-paced environment. Willing to learn and can-do attitude
• Master’s/Bachelor’s/PhD degree in Statistics, Economics, Finance, Mathematics or a related quantitative field
• Engineering/MBA background from a premier institute

Specialization in fraud or risk domain preferred