Skip to contentSkip to footer
A global semiconductor manufacturer was focused on helping its clients meet the moment as they transform the power grid and build out the digital world. But that demand presented a challenge to manage across many international markets. The company knew that automating demand forecasting as a key component of its production and logistics planning could allow it to optimize its operations. That's where we came in.
With approximately 30,000 items to order and 100 supply chain professionals regularly adjusting forecasts, the company aimed to reduce manual labor as much as possible without compromising the accuracy of the forecasts. Our AI and Industrials teams were brought on to develop a methodology for assessing and improving the forecasting process.
Short- and long-term forecasts for growth
The development of a forecasting model for demand with a time horizon of six months was based on internal historical data using state-of-the-art machine-learning techniques. In addition, a strategic long-term demand forecast was created for up to 18 months, taking into account both external and internal data.
The automated forecasts proved to be precise and robust. Following the successful transfer of knowledge, the company was able to further develop the solution independently. It is now an integral part of S&OP planning.
The implementation resulted in the automation of 80% of product demand forecasts and a 75% reduction in manual effort, without compromising the accuracy of the forecasts.
80%
forecasts automated
75%
reduction of the manual workload
