Safeguarding food supply with AI-enhanced demand forecasting

During the COVID-19 pandemic, a Fortune 500 consumer packaged goods (CPG) company, faced unprecedented challenges in demand forecasting due to rapidly shifting consumer behaviors. Initially seeking a solution for an AI project, the company’s engagement with Inspire11 quickly expanded as we identified critical issues within their forecasting models and broader data science practices.

SOLUTION

Development of Demand Forecasting Tool

Created a robust demand forecasting tool designed to incorporate external datasets such as industry trade data, weather patterns, Google trends, and CDC data. This enriched model provided deeper insights into forecast drivers and allowed for comprehensive scenario analysis through custom dashboards.

Machine Learning Operations (MLOps) Pipelines

Implemented MLOps pipelines to enhance collaboration between IT and the demand science team. This empowered the team to independently generate and refine insights, accelerating issue resolution and operational efficiency.

Generative AI for Explainable Forecasting

Integrated generative AI to analyze market outlooks, compare them to actual performance, and explain forecast variations. This innovative approach allowed our client to identify gaps and drive their data-driven success to new heights.

IMPACT & OUTCOMES

Enhanced Forecast Accuracy: Achieved a 1.5% variance in weekly forecasts for individual products, enabling more frequent updates and the ability to run multiple scenarios incorporating pandemic-related drivers.

Operational Agility: Improved the collaboration between business and IT through MLOps pipelines, allowing the demand science team to operate more efficiently and independently.

Strategic Insights: The integration of generative AI provided explainable forecasting, which helped our client better understand market dynamics and adjust strategies accordingly.