Anticipating demand accurately is key to driving decisions in any supply chain. With sales cycles shortening, consumer preferences fluctuating, and millions of data points generated every second, traditional forecasting techniques are not sufficient for today’s markets. Aleph OSC’s proprietary demand sensing algorithms combine your demand planning estimates, traditional forecasting and exogenous variables, fed by big data, to enable Aleph OSC’s preemptive ordering capabilities.

We use contemporary machine learning and model ensembles to obtain precise demand estimates by location and SKU. Our platform automatically selects the best combination of models and datasets to improve demand prediction accuracy for each location. Additionally, thanks to being built on top of distributed computing engine Apache Spark, data sources can be as large as necessary.

In addition to demand prediction, our platform can also learn from historical data to estimate lead times and costs for every route. This influences Aleph OSC’s decision making by adapting to changing networks across time. If certain events throughout the year make lead times significantly increase in the morning, Aleph OSC can adapt by rescheduling deliveries to the best time of day.