S
Senior Data Scientist, Recommendations and Personalization
Not specifiedFull-timeData ScienceEntertainment
Own applied machine learning for recommendations and personalization, from behavioral data and model adaptation through production deployment and impact measurement.
Responsibilities
- Own recommendation and personalization across the user journey and product surfaces
- Build predictive models for retention, churn, conversion propensity, and lifetime value
- Adapt existing models and architectures to the company’s data
- Deploy and own models in production, including serving, monitoring, retraining, and iteration
- Design and analyze experiments to measure business impact
- Partner with Product, Content, and Growth teams to apply model outputs
- Replace manually tuned heuristics with learned models
Requirements
- BSc in a quantitative field such as data science, computer science, statistics, or mathematics
- At least 5 years of data science or machine learning experience, including production model deployment
- Hands-on production experience building recommendation or personalization systems
- Strong applied machine learning skills, including modeling, evaluation, and feature engineering
- Hands-on experience training and adapting neural networks and deep learning models for production
- Experience modeling time series, including forecasting, trends, seasonality, and noisy data
- Ability to adapt existing models or architectures to real-world data
- Experience deploying and maintaining production models
- Knowledge of A/B testing, causal reasoning, and measurement
- SQL skills and experience working with large data warehouses
Nice to have
- MSc or PhD
- Experience with recommendation systems at scale, including retrieval, ranking, sequence or session models, and embeddings
- Consumer or subscription experience with LTV, churn, or propensity modeling
- Streaming, entertainment, consumer mobile, or growth-marketing data experience
- MLOps experience with feature stores, serving, monitoring, or retraining pipelines
- Causal inference or uplift modeling