Case studies

Given the multitude of factors involved, demand forecasting is too complex to be performed manually. The company therefore partnered with Vooban to develop an AI-powered solution to predict daily demand for each of the company’s three key processes: parcel pickup, sorting, and delivery.

To improve the overall efficiency of Nationex’s supply chain, Vooban developed two AI models that provide forecasts across different time horizons.

The short-term model, covering a 7-day window, helps identify key levers to reduce operational costs for the upcoming week. The long-term model, spanning the next 12 months, serves as a decision-support tool for the company’s budgeting process. The primary goal is to predict the daily volume of parcels to optimize operations on the road, at sorting centers, and in depots, while also determining staffing requirements. To support both models, a simulation and visualization tool was created, enabling teams to explore various hypothetical scenarios that could impact planning, such as onboarding a new client or adjusting resource allocation.

Starting from an initial average accuracy of less than 60%, the demand prediction models now achieve 80% accuracy at the six-month mark and up to 89% just one week in advance thanks to the models developed. This level of precision enables much more refined planning, improves the overall supply chain efficiency and contributes to tangible reductions in operational costs, all while minimizing environmental impact.

In one of the most successful instances, the model predicted a volume of 50,300 parcels on a day where over 50,000 parcels were actually processed—demonstrating an impressive ability to anticipate demand even under complex conditions.

By improving both productivity and service quality, Nationex ensures it remains competitive in a market dominated by major players. Ultimately, an optimized supply chain allows them to take on more clients while keeping operational costs low, paving the way for rapid growth.