CAREERS
    CONTACT US
    sbi_logo_positif240x240px_cropped
    Back to achievements

    Optimizing Financial Management of Raw Materials with Data Science and Automation

    • Practice
    Data Platform Data Visualization AI
    • Team

    1 Architect
    1 Data engineer
    2 Data scientists
    2 Business experts

    • Technical environment
    €10M
    Annual projected savings
    €7M
    Achieved in 7 months
    +10 use cases
    Scalable model expansion
    CHALLENGES

    Danone, a global leader in food and beverages, operates across four strategic divisions: fresh and plant-based dairy products, bottled water, medical nutrition, and infant nutrition. Listed on Euronext Paris and part of the CAC 40 index, Danone has been the world’s first listed "entreprise à mission" (purpose-driven company) since 2020, committed to bringing health through food to as many people as possible.

    The company faces a critical challenge: the volatility of raw material costs, particularly plastic (including polyethylene and PET), with annual procurement reaching €700–800 million.

    Key Issues:

    • Plastic prices are highly volatile, directly tied to oil markets, and significantly impact P&L and product margins.
    • The current decision-making process is manual, slow, and inefficient.
    • Objective: Improve procurement and hedging strategies, with an estimated €10M annual gain.
    SOLUTION
    With SBI's help, Danone combined data science and robotic process automation (RPA) to:
    • Automate data collection from over 20 market reference sources for real-time monitoring.
    • Standardize and clean data to ensure consistency and compliance.
    • Build a dedicated data warehouse for historical and predictive pricing analysis.
    • Develop predictive algorithms (including XGBoost) to forecast price trends.
    • Validate the approach with procurement and finance teams to ensure buy-in.
    • Design custom dashboards tailored to operational needs and business culture.
    BENEFITS
    • Enhanced visibility: Real-time, data-driven insights for strategic procurement decisions.
    • Scalable model: Easily adaptable to new PET sources and other raw materials.
    • High predictive accuracy: Powered by machine learning and advanced algorithms.
    • Unified data hub: Centralized historical database for reliable price tracking and analysis.
    • Actionable dashboards: Clear, shared, and standardized view of "price truth."
    • Immediate financial impact: €7M saved in 7 months, with a €10M annual potential.

    This initiative not only delivered proven results for plastic procurement but also unlocked +10 new use cases for other raw material categories, reinforcing Danone’s operational resilience and competitiveness.

    Why This Matters

    By leveraging data science and automation, Danone transformed a reactive procurement process into a predictive, agile, and cost-efficient model—setting a benchmark for the industry.