Global Agri Major set out to forecast annual yields for multiple crops across North America at both macro and micro levels. The approach rolled country-level views into a full North American forecast to support planning and market positioning. The engagement started in January 2019.
The team needed a dependable way to bring together diverse agricultural signals – ranging from weather and satellite data to supply-and-demand context and market sentiment – and turn them into timely, usable predictions across regions and crops.
Using the NorthGravity Platform, the project aggregated comprehensive weather datasets and satellite images, combined them with supply-and-demand metrics, and incorporated news and social media insights to capture real-time sentiment. Client machine-learning models built in Python and R processed these inputs and produced more than 10,000 predictions each week, with production results delivered in 30 days.
The organization obtained accurate yield forecasts for North America ahead of government publications, improving readiness at both country and regional scales and enabling data-driven action earlier in the season.
