(USA) Senior Manager, Data Science - AI Agent Development

About the role

Position Summary... What you'll do... Model Assessment and Validation: Requires knowledge of model fit, testing, tuning, and validation techniques (e.g., Chi-square, ROC curve, root mean square error, etc.), impact of variables and features on model performance, to identify the model evaluation metrics. Apply best practice techniques for model testing and tuning to assess accuracy, fit, validity, and robustness for multistage models and model ensembles. Data Visualization: Requires knowledge of visualization guidelines and best practices for complex data types; multiple data visualization tools (for example, Python, R libraries, GGplot, Matplotlib, Plotly, Tableau, PowerBI, etc.); advanced visualization techniques/tools; multiple story plots and structures; OABCDE communication, influencing technique, emotional intelligence. To generate appropriate graphical representations of data and model outcomes, understand customer requirements to design appropriate data representation for multiple data sets. Work with User Experience designers and User Interface engineers as required to build front-end applications. Present to and influence the team and business audience using the appropriate data visualization frameworks and convey clear messages through business and stakeholder understanding. Customize communication style based on stakeholder (under guidance) and leverage rational arguments. Guide and mentor junior associates on story types, structures, and techniques based on context. Understanding Business Context: Requires knowledge of industry and environmental factors, common business vernacular, business practices across two or more domains (such as product, finance, marketing, sales, technology, business systems, and human resources), and in-depth knowledge of related practices, directly relevant business metrics, and business areas. To provide recommendations to business stakeholders to solve complex business issues, develop business cases for projects with a projected return on investment or cost savings, translate business requirements into projects, activities, and tasks, and align to overall business strategy and develop domain-specific artifacts. Serve as an interpreter and conduit to connect business needs with tangible solutions and results. Identify and recommend relevant business insights pertaining to their area of work. Analytical Modeling: Requires knowledge of feature relevance and selection, exploratory data analysis methods and techniques, advanced statistical methods and best-practice advanced modelling techniques (e.g., graphical models, Bayesian inference, basic level of NLP, vision, neural networks, SVM, Random Forest, etc.), multivariate calculus, statistical models behind standard ML models, advanced Excel techniques, and programming languages like R/Python, basic classical optimization techniques (e.g., Newton-Raphson methods, gradient descent), numerical methods of optimization (e.g., linear programming, integer programming, quadratic programming, etc.). To select appropriate modeling techniques for complex problems with large-scale, multiple structured and unstructured data sets, select and develop variables and features iteratively based on model responses in collaboration with the business. Conduct exploratory data analysis activities (for example, basic statistical analysis, hypothesis testing, statistical inferences) on available data. Identify dimensions and designs of experiments and create test-and-learn frameworks. Interpret data to identify trends to go across future data sets. Create continuous online model learning along with iterative model enhancements. Develop newer techniques (for example, advanced machine learning algorithms, auto ML) by leveraging the latest trends in machine learning/artificial intelligence to train algorithms to apply models to new data sets. Guide the team on feature engineering, experimentation, and advanced modeling techniques to be used for complex problems with unstructured and multiple data sets (for example, streaming data, raw text data). Model Deployment and Scaling: Requires knowledge of impact of variables and features on model performance, understanding of servers, model formats to store models. To deploy models to production, continuously log and track model behavior once it is deployed against the defined metrics. Identify model parameters which may need modifications depending on scale of deployment. Code Development and Testing: Requires knowledge of coding languages like SQL, Java, C, Python, and others; testing methods such as static, dynamic, software composition analysis, manual penetration testing, and others; business domain understanding. To write code to develop the required solution and application features by determining the appropriate programming language and leveraging business, technical, and data requirements. Create test cases to review and validate the proposed solution design. Create proofs of concept. Test