Demand Forecasting Engine
Knowing what tomorrow's customers want
The core of all OPAL solutions is the Demand Forecasting Engine (DFE). This generic software component is based on Artificial Intelligence methods. The DFE automatically learns from historical data and provides very close predictions about the expected demand. Patterns that result from calendar (weekdays, holidays) and seasonal (season, weather) constellations and spontaneous events (entertainment events, thunderstorms) are projected onto the future.
As you might know from weather, forecasts are subject to uncertainty. For this reason, the DFE takes cost-optimal safety stock intervals into account in its forecasts. Likewise, the DFE also computes cost-optimal lot size rounding.
How does DFE work?
The DFE is a generic software component that automatically links your historical data, such as sales and losses, and machine data, with external data, such as weather and calendar constellations. From this input, DFE calculates precise forecasts of future demand.
In order to identify the complex relationships in the data, machine learning (ML) techniques, also known as Artificial Intelligence (AI), are used. For example, a genetic algorithm automatically optimizes the linking of your data with relevant, external features. It also chooses the appropriate time series models to have the most accurate predictions possible. The time series models are based on artificial neural networks (ANN), gradient boost models (GBM) or generalized linear models (GLM).
And that's not all. In order to cost-effectively handle the existing uncertainty, the DFE calculates the optimal assortment, the optimual shelf display for the end of a day, and the optimal lot sizes for the delivery and production on site.