Senior Data Scientist (Retention & Product)

Join CookUnity as a Senior Data Scientist focused on retention and churn. In this hands-on role, you will predict at-risk customers, understand why they leave, and power interventions to keep them ordering. You will work closely with Product, CRM, Marketing, and Engineering teams. Responsibilities include building churn/survival models, powering retention interventions, developing propensity models for win-back, and optimizing offers and promotions. You will also own the full model lifecycle and design experiments for uplift measurement.

• Build churn/survival models and lifecycle-state models that flag at-risk customers early and anticipate where each customer is heading.
• Power the retention interventions that act on that risk — save flows, skip/pause deflection, lifecycle messaging — and classify why customers churn.
• Develop propensity models and personalized experiences that bring churned customers back.
• Retention/churn depth: churn prediction, survival / time-to-event modeling, and lifecycle-state models in a subscription or recurring-revenue context
• 5-8+ years in data science, applied ML, or statistics, shipping production models
• End-to-end ML & MLOps: building, validating, and deploying models in production (CI/CD, registries, containerization, orchestration, monitoring)
• Collaboration: excellent communication; able to embed with Product/CRM/Marketing and turn models into decisions
• Causal & experimentation rigor: uplift/incrementality measurement and A/B testing, with the judgment to separate true impact from selection effects
• Engineering & tooling: strong Python (pandas, scikit-learn, gradient boosting; deep learning a plus), SQL, code hygiene and reproducibility
• Ability to leverage generative AI to increase output quality and speed
• Education: BS in a quantitative field required; MS/PhD preferred
• Recommenders, embeddings, or personalization for retention and win-back
• Subscription marketplaces, food-tech, or consumer marketplaces with a retention mandate
• Lifecycle-state / Hidden Markov models for churn
• Causal and uplift libraries (e.g. EconML) or survival-modeling packages