Data Scientist (Churn/Retention/Propensity)

Project overview

Contract Data Scientist for a hybrid (3x onsite) role in Santo Amaro, São Paulo, starting immediately. The focus is building, validating, deploying, and monitoring predictive ML models (churn, retention, propensity) using Python and SQL on large/complex datasets. The role requires hands-on modeling with Random Forest and XGBoost, model interpretability (e.g., SHAP/feature importance), strong documentation, and the ability to translate analytical results into business actions. Cloud experience and cloud infrastructure knowledge are required; AWS/Azure/Oracle Cloud, Spark/Databricks, SAS, Tableau are desirable.

Advanced Excel is required.

Responsibilities

• Develop, validate, and evolve predictive models of churn, retention, and propensity.
• Evaluate and explain the behavior of models using interpretability techniques, such as SHAP, Feature Importance, among others.
• Translate analytical results and recommendations into strategic actions for the business areas.
• Perform analysis and propose improvements in the performance of the models through new variables, algorithms and modeling strategies.

Requirements

• Python
• SQL
• Machine Learning (model development/validation/monitoring)
• Random Forest
• XGBoost
• Churn/retention/propensity modeling
• Model interpretability (e.g., SHAP, feature importance)
• Working with large volumes of data and complex databases
• Production implementation and monitoring of ML models
• Cloud experience
• Cloud infrastructure knowledge
• Documentation of analytical solutions
• Advanced Excel
• Ability to translate business problems into data-driven solutions
• Communication/presentation of results
• Collaboration/teamwork

Project details

• Start: flexible
• Remote: hybrid
• Location: Brazil, BR