Data Science for Supply Chain Forecasting
4.5
Read more about Data Science for Supply Chain Forecasting on Amazon
SupChains
We build demand forecasting and inventory planning models for manufacturers, distributors and retailers. Two subjects, nothing else — and one promise we measure: forecast error down by up to 30%, inventory down by up to 20%.
Let’s talk
About SupChains
We are a specialized team on demand planning and inventory planning. Our clients come to us because they want a deep expertise on supply chain planning to help them improve their demand planning process, deploy machine-learning forecasting models, and ultimately beat their accuracy targets.
We are not implementors, we are not consultants. Instead, we build the models and deliver forecasts for our clients as a service. We developed our own SupChains App so planners can not only review and enrich forecasts, but leadership can also track Forecast Value Added across all dimensions.
Our App is built to support the SupChains Way of doing demand planning. Find more about it here.
Who is behind SupChains
SupChains was founded in Brussels in 2016 by Nicolas Vandeput.
He wrote the three books our method comes from. He teaches demand forecasting and inventory optimization to master students at CentraleSupélec in Paris since 2020, at Albert School since 2024 and at Mines Paris since 2025.
He runs VN1 and VN2, two international forecasting and inventory competitions, and writes on Medium.
The books
Our founder wrote three books on demand forecasting and inventory optimization, with around 15,000 copies sold. They are where our method comes from, and each one ships with the datasets and the models, so you can run them yourself.
Inventory Optimization: Models and Simulations
4.7
Read more about Inventory Optimization: Models and Simulations on Amazon
Demand Forecasting Best Practices
4.8
Read more about Demand Forecasting Best Practices on Amazon
What we can do for your team
A demand forecasting model
One machine-learning engine across your whole portfolio, benchmarked against your current process. We start with a proof of concept; after that it runs as a service or under licence.
An inventory policy optimizer
Simulation instead of the safety-stock formula, tuned per item. Inventory down by up to 20% at the same service level.
An audit
Three to five weeks. Where your forecast really loses accuracy, and which change would actually help.
Training, coaching and workshops
Live online sessions, two hours at a time, for your planners — or coaching for the people who lead them.
The SupChains Way
Over the years, we developed our own way of doing demand and inventory planning. It is what we apply on every client project, and it is how the winning teams of the VN1 and VN2 competitions work too.
At its core: let a robust, global machine-learning engine do the mechanical forecasting work, ask planners to bring the information the engine cannot see, measure the value added at every step, and optimize inventory policies by simulation rather than trusting a formula.
Download the SupChains WayMaster demand and inventory planning
Join a community of 10,000+ professionals achieving supply chain excellence.
Every week in your inbox: practical insights on demand forecasting and inventory optimization, expert opinions, and invitations to exclusive webinars & masterclasses.
You will also receive 4 chapters of Nicolas Vandeput’s book Demand Forecasting Best Practices — absolutely free.
Get the free chapters