Machine Learning Engineer, Quizlet

• 6+ years of applied ML or ML-heavy engineering experience, with a track record of shipping production models that drive measurable business impact
• Deep expertise in classical ML techniques (e.g., boosted trees, GLMs, survival models, uplift modeling)
• Experience with reinforcement learning, contextual bandits, or sequential decision-making
• Strong engineering skills with Python and common ML frameworks (scikit-learn, PyTorch, XGBoost, LightGBM, etc.)
• Demonstrated experience integrating ML systems into complex product architectures, ideally including monolithic applications
• Experience defining integration boundaries, solving backend/ML interface issues, and collaborating with infra teams on serving patterns
• Strong understanding of experimentation design, causal analysis, and the relationship between offline and online evaluation
• Excellent communication skills for conveying technical constraints and integration trade-offs
• A strong ownership mindset centered on reliability, maintainability, and long-term system health

Desirable

• We are looking for Machine Learning Engineers ranging from the Senior to Staff levels (note: leveling decisions made through the interview process)
• Within this organization, this role is responsible for the predictive and decisioning models that drive monetization, retention, activation and goal-aligned study guidance
• These systems balance immediate impact with long-term user value and must integrate seamlessly into Quizlet’s product architecture
• You will lead both the modeling efforts and the technical integration work required to bring complex ML systems into production
• This includes designing predictive and prescriptive models (such as conversion propensity, churn risk, LTV, sequential decisioning, and timing optimization) and collaborating closely with product and infrastructure engineering to ensure these models can be safely and cleanly embedded into existing product workflows
• A major part of this role involves identifying dependencies within the product codebase, defining integration contracts with cross-functional partners, and shaping technical solutions that allow ML-driven decisioning to operate reliably, efficiently, and maintainably at scale
• You’ll work closely with product managers, data scientists, platform engineers, backend engineers, and fellow ML engineers to deliver ML-driven experiences that drive engagement, satisfaction, and measurable business outcomes
• Lead the design and development of predictive and prescriptive models (e.g., conversion propensity, churn risk, LTV, uplift, sequential decisioning, and timing optimization) that drive learner-facing decisions across monetization, lifecycle, and study guidance surfaces
• Design and build decisioning and policy models that determine learner-facing actions across product surfaces, including monetization, lifecycle, and study guidance use cases. These systems operate under real-world product constraints and must optimize across multiple, sometimes competing objectives
• Determine when and how to present paywalls, discounts, or value exchanges
• Selecting personalized study modes or interventions based on learner state, intent, and context
• Triggering retention and churn-prevention actions at the appropriate moment
• Balancing short-term conversion and revenue goals with long-term engagement, retention, and learning outcomes
• Prioritize: Multi-objective optimization across monetization, retention, user experience, and learning outcomes, time-aware and eligibility-aware decisioning, rather than static prediction, consistent action selection across sessions, devices, and product surfaces, and an approach that connects offline modeling metrics to online experimental results
• Apply and advance uplift modeling, survival analysis, sequential decisioning, and other policy-based approaches, taking responsibility for bringing these techniques into production-grade systems
• Lead the end-to-end productionization of ML systems, from modeling through integration, ensuring models can be safely, cleanly, and reliably embedded into existing product workflows
• Identify upstream and downstream dependencies within the product codebase and data ecosystem, and proactively address integration risks
• Define and negotiate clean integration boundaries, including API contracts, data interfaces, decision schemas, and fallback strategies, in collaboration with product and infrastructure engineering
• Partner closely with Infrastructure Engineering to design scalable, resilient, and observable model-serving paths that integrate with Quizlet’s application stack
• Embed model-driven decisioning logic into backend and product flows in ways that are maintainable, testable, and compatible with existing systems
• Build and maintain end-to-end pipelines for feature engineering, training, evaluation, deployment, and monitoring, ensuring training–serving consistency
• Improve latency, throughput, reliability, and observability of real-time and near–real-time inference systems operating at scale
• Translate product goals (conversion, retention, revenue, engagement) into clear modeling objectives and technical specifications
• Collaborate closely with product managers, backend engineers, and infrastructure partners to ensure ML systems fit naturally into the existing architecture without introducing brittle dependencies
• Develop evaluation frameworks that tie offline metrics to online A/B results, ensuring changes are measurable, interpretable, and aligned with product impact
• Clearly communicate assumptions, trade-offs, risks, and technical constraints to both technical and non-technical stakeholders
• Provide technical leadership for ML-driven decision systems, guiding the organization toward unified policy models and consistent action-selection frameworks across surfaces
• Mentor engineers and scientists, setting a high bar for modeling rigor, production quality, experimentation discipline, and responsible ML
• Shape long-term strategy for scalable, maintainable ML decisioning, bringing modern approaches—including sequential decisioning and RL-adjacent techniques—into production where appropriate