Data Scientist

Caterpillar · Industrial · Piracicaba, São Paulo

Data Scientist role focused on transforming data into advanced analytical and predictive insights for decision-making within Finance. This role involves data exploration, statistical and machine learning modeling, data preparation, and translating analytical outputs into actionable insights. It supports automation and advanced analytics enablement, model monitoring, and business partnership, working closely with IT and Data Platforms.

What you'd actually do

  1. Explore structured and semi‑structured data to understand business behavior, identify patterns, outliers, and relationships, and generate hypotheses.
  2. Develop, test, and support statistical, predictive, or machine learning models (under guidance for junior levels) to support forecasting, risk analysis, and decision‑making.
  3. Prepare datasets through cleansing, transformation, and feature creation to ensure analytical reliability and usability.
  4. Translate analytical outputs into clear, actionable insights, explaining assumptions, limitations, and business implications to stakeholders.
  5. Support automation and data‑driven solutions by embedding analytics into business processes, dashboards, or Power Platform assets.

Skills

Required

  • University degree
  • English Intermediary
  • Knowledge in SQL Server (for querying, data manipulation and complex joins)
  • Knowledge with Python for data analysis
  • Advanced Microsoft O365 (Power Platform - Power BI, Apps, Automate)

Nice to have

  • Advanced Experience with Python for data analysis (e.g., pandas, numpy, matplotlib, scikit-learn)
  • Advanced SQL (CTEs, performance tuning, complex joins)
  • Data modeling concepts
  • Statistics and basic machine learning techniques
  • ETL / data preparation processes
  • Exposure to Agile or project management methodologies (Scrum, Kanban, Lean, Black Belt)

Other signals

  • Develop, test, and support statistical, predictive, or machine learning models
  • Explore structured and semi-structured data to understand business behavior, identify patterns, outliers, and relationships, and generate hypotheses
  • Prepare datasets through cleansing, transformation, and feature creation to ensure analytical reliability and usability