How an AI-driven demand forecaster works | A talk with an IG&H developer

IG&H Quantivate is a solution suite for retail professionals, built in-house by our data scientists, UX/UI designers and developers. We spoke with the lead developer, Niels de Brabander about how an AI-driven demand planning solution works.

Why should retailers use Artificial Intelligence?

With the retail industry becoming increasingly complex and interlinked, COVID was a clear example of how important it is to be able to handle disruption. Rule-based systems aren't accurate enough anymore because they function best for predictable trends.

Technology like Machine Learning and Artificial Intelligence help you react proactively. For instance, the Quantivate system is updated daily and reconsiders every new data point.

What is Quantivate?

Quantivate is a platform hosting multiple solutions that entail everything related to demand planning. Quantivate has a demand forecaster, DC replenisher and a store replenisher. These modules build upon each other. We started with the demand forecaster because it serves as the basis for a lot of the other modules.

How did Quantivate originate?

I was first building bespoke forecasting solutions, tailored to a specific client’s needs. With that experience, I started to work on Quantivate together with Erwin van Beek, one of our data engineers. I joined the Quantivate project at the start of 2021 and we have been building the solution since then. I lead the development team and am enjoying the process.

Where is the AI in Quantivate?

In the data science field, we usually talk about machine learning models or algorithms. The AI and Machine Learning is part of the platform’s code.

What makes Quantivate different from other demand planning systems?

It’s based on AI, not on business rules. This means it not only allows better prediction but is also easier to implement. Users don’t need to create a lot of rules just to get the system up and running.

The underlying vision of Quantivate is that we can implement a system that needs very little user input. Since all it requires is a data feed, this also means a shorter implementation time.

What does implementation of Quantivate look like?

The main task of implementation is getting the data and curating it so it can be processed by system. It's a matter of weeks, which is comparatively fast! With other solutions, IT Consultants are busy with implementation for several months. Therefore, Quantivate's implementation is quicker and the TCO (total cost of ownership) is lower. During implementation, we follow three main steps:

1) Set up the data feed

2) Ensure the retailer is connected in terms of sending their data to the platform

3) Match the retailer's and the platform's definitions

It's a cloud-based platform which means there's no physical installation required on the client's side.

IG&H Quantivate is a solution suite for retail professionals. With the use of AI, you benefit from decision support that not only identifies but also makes connections that the human mind may not be able to. With optimal demand prediction, you are better equipped to handle complex and unpredictable scenarios.

Would you like to find out more about this retail solution?