Machine Learning Engineer, Causal Inference, Level 5

Overview As a Machine Learning Engineer at Snap, you design causal models and productionize ML solutions to uplift user, advertiser, and business value. You’ll work with product and engineering teams to shape experimentation strategies and interpret results. You’ll balance model complexity, bias, scalability, and interpretability while upholding rigorous engineering standards. This role offers impact in a fast-moving, privacy-conscious environment that aims to advance how people live and communicate. Compensation / Benefitspaid parental leavecomprehensive medical coverageemotional and mental health supportequity in RSUscompensation packagelong-term success sharing ResponsibilitiesDesign and build causal models that quantify impact and optimize decision-makingProductionize causal ML solutions (e.g., uplift modeling, heterogeneous treatment effect estimation) using observational and experimental dataDesign, analyze, and interpret A/B tests and quasi-experiments; collaborate with product and engineering partners on experimentation strategiesEvaluate tradeoffs among model complexity, bias/variance, scalability, and interpretabilityConduct code reviews and maintain scalable, maintainable infrastructureContribute to rapid iteration cycles while maintaining methodological rigor Key requirementsBachelor’s degree in computer science, statistics, economics, or a related technical field, or equivalent practical experience5+ years of post-Bachelor’s experience in machine learning with hands-on causal inference or experimentation; or Master’s + 4+ years; or PhD + 2 yearsDemonstrated experience building models to support product decision-making and policy evaluation through causal techniquesExperience designing and analyzing online experiments (A/B tests) and leveraging causal ML in production systemsStrong understanding of causal inference and modern methods (meta-learners, propensity score matching, instrumental variables)Proficiency in Python and libraries (pandas, NumPy, scikit-learn, CausalML, etc.)Strong communication and ability to translate technical insights for non-technical partnersAbility to work independently and collaborate across cross-functional teamsstrong communication and mentorshipcollaboration across cross-functional teamsability to translate technical concepts for non-technical audiencescausal inference and treatment effect estimationuplift modeling and experimentation infrastructureA/B testing and quasi-experiments