Job Description

This role supports end-to-end model lifecycle management. Responsibilities encompass model development, independent validation and assessment, performance optimization, monitoring, documentation, and governance — ensuring all models adhere to applicable standards and remain fit-for-purpose throughout their operational life.

You will work within a cross-functional team to build propensity models, next-best-action (NBA) engines, customer segmentation frameworks, and campaign response models that power data-driven marketing strategies. You will apply best practices in model explainability, fairness testing, and lifecycle governance to ensure high-quality, compliant model outputs.

• Develop and optimize propensity models, segmentation frameworks, and campaign response models using supervised and unsupervised machine learning techniques.
• Conduct bias testing and implement model explainability methods (e.g., SHAP, LIME) to support model approval and governance workflows.
• Perform exploratory data analysis and feature engineering to identify meaningful predictors of customer behavior.
• Support the full model lifecycle including development, documentation, validation support, performance monitoring, and periodic re-validation.
• Collaborate with marketing and data teams to understand business objectives and translate them into modeling requirements.
• Prepare clear, concise model documentation including methodology overviews, performance summaries, and limitation disclosures.
• Monitor deployed models for data drift, performance degradation, and champion-challenger evaluation.
• Contribute to A/B test design and campaign measurement analyses to evaluate marketing effectiveness.
• Stay current on advances in marketing data science, uplift modeling, and customer analytics.

• 3–6 years of experience in data science, analytics, or quantitative modeling with a focus on customer or marketing analytics.
• Hands-on experience building and evaluating classification and regression models for propensity scoring or segmentation.
• Proficiency in Python with working knowledge of libraries such as scikit-learn, XGBoost, LightGBM, pandas, and numpy.
• Familiarity with model explainability tools (SHAP, LIME) and basic concepts of model fairness and bias testing.
• Experience with SQL and large-scale data platforms for data extraction and feature preparation.
• Understanding of model validation principles and documentation standards.
• Strong analytical and problem-solving skills with attention to detail.
• Effective written and verbal communication skills for presenting model results to both technical and business stakeholders.
• Bachelor's degree (Master's preferred) in Statistics, Mathematics, Computer Science, Engineering, or a related quantitative field.