Client Industry
Argriculture
Technology stack
Computer Vision
Machine Learning
Industry
Agriculture
Technology stack
Computer Vision
Machine Learning
Industry
Agriculture
Technology stack
Computer Vision
Intelligent Automation
The challenge
Their existing prediction process relied on older on-premise systems and deep individual staff knowledge. The process was difficult to maintain, slow to update, and created operational risk as experienced employees approached departure. Forecast accuracy and system reliability were becoming concerns.
The solution
Rather than redesigning everything, the focus was on stabilising and modernising the existing process. The prediction models and data were moved into a central data platform (Databricks) that is easier to manage, monitor, and maintain.
The transition followed a minimal change approach to keep the logic and outputs familiar for users while removing manual steps, reducing reliance on individual staff, and improving system reliability.
The outcomes
- More consistent and reliable marbling and carcass weight forecasts
- Reduced dependence on key individuals to run and maintain the process
- Less manual effort required to manage predictions
- Improved confidence for sales and finance teams when planning and pricing
- A stable foundation for future improvements without operational disruption
