Data Scientist

• Degree in Data Science, Computer Science, Engineering, Applied Mathematics, or a related field.
• At least 1 year of commercial experience as a Data Scientist or Machine Learning Engineer.
• Proficiency in Python and core data and machine learning libraries, including pandas, NumPy, scikit-learn, and statsmodels.
• Hands-on experience with gradient boosting models.
• Solid understanding of time-series forecasting, regression modeling, and cross-validation techniques.
• Knowledge of regression evaluation metrics in business contexts, such as MAE, R-squared, and NRMSE.
• Experience with version control systems such as Git and clean, modular coding practices using OOP.
• Proficiency in SQL.
• Comfortable working in a Linux environment.
• Ability to present and communicate results to technical and business stakeholders, translating model outcomes into clear business value.
• Communicative English, sufficient for reading documentation and presenting results.

Nice to have:

• Familiarity with CPG/FMCG or Retail data and domain-specific metrics.
• Experience with cloud environments, preferably AWS.
• Practical use of Explainable AI tools, such as SHAP.
• Experience with hyperparameter optimization frameworks, such as Optuna.
• Working knowledge of Excel for ad-hoc data analysis and reporting.

Predictive Modeling:

• Build and improve sales volume forecasting models using gradient boosting methods and regression techniques, including hyperparameter optimization.

Data Analysis & Feature Engineering:

• Explore and prepare data for modeling, including time-series feature creation, promotion and pricing analysis, outlier detection, and data quality assurance.

MLOps:

• Operationalize models and work within a production-grade machine learning pipeline leveraging cloud infrastructure.

Explainable AI (XAI):

• Collaborate with business teams to interpret and communicate model outputs, translating analytical results into actionable business insights.

Research:

• Test new hypotheses, evaluate alternative loss functions and algorithms, and explore improvements to existing modeling approaches.
• Pursue independent research topics, with the opportunity to publish findings of novel scientific value.

• Fully remote work includes on-site collaboration for 2–3 days per quarter.