Data Scientist

Lingaro is a company operating since 2008 with over 1500 talents across 7 global sites.

• Commercial experience with various classical data science and machine learning models (e.g., decision trees, ensemble-based tree models, linear regression).
• Knowledge of customer analytics concepts or advanced forecasting.
• Model hyperparameter tuning.
• Fluency in Python.
• Basic working knowledge of SQL.
• Knowledge of specific data science/machine learning libraries.
• Solid experience in one of the cloud computing platforms (Databricks, GCP, or Azure).
• Basic computer programming skills and familiarity with programming concepts.
• Strong business acumen.
• Experience with deep learning, reinforcement learning, or other advanced modeling concepts in classical data science problems.

Nice to have:

• Understanding of causal machine learning.
• Experience working with big data and distributed environments.
• Commercial experience proven by multiple successful projects in forecasting.
• Experience with object-oriented programming (OOP) in Python.
• Experience with MLOps.
• Familiarity with other languages such as R and Scala.

• Work on end-to-end classification and forecasting use cases: problem framing, data preparation, model development, evaluation, and basic deployment support (e.g., demand forecasting, churn prediction).
• Explore and clean data; perform exploratory data analysis (EDA) to understand data and flag data quality issues.
• Engineer features for tabular and timeseries data.
• Train, validate, and tune standard machine learning models (e.g., logistic regression, tree-based models, gradient boosting, simple neural nets, classical timeseries models).
• Evaluate models with appropriate metrics that impact business KPIs.
• Build clear visualizations and concise reports to present model results and insights to business stakeholders.
• Collaborate with data engineers and AI engineers to bring models into production (batch scoring, APIs, models monitoring, dashboards).
• Document data sources, modeling assumptions, and experiment results in a reproducible way (notebooks, reports, wikis).