The following is an excerpt from the Designed Analytics Report, “Generative AI in Supply Chain: Key Considerations“, publishing on 02/25/2024.

If we start listing out the potential opportunities of Generative AI in supply chain, or for that matter any other function, the list can be long. However, like every other digital transformation opportunity, there is a need to carefully evaluate the impact, vs the resources that need to be invested to build certain capabilities.

LLM training is much more resource intensive than some other ML based approaches that organizations are currently leveraging. What this means is that there may be scenarios where even though the value delivered may be only slightly more as compared to another use case, the training and other requirements, to help make the vision a reality, will be more or less the same as other Generative AI based capabilities. This means that developing a roadmap, based on practical evaluation of the capabilities you want to develop, becomes important.

The first step is to understand that Generative AI capabilities across supply chain, and many other functions, can be categorized into three broad categories. These three categories are productive, insightful and strategic.

The capability pyramid shown in Figure 4 is not supply chain specific. You can extrapolate this to any other function. But the key message here is to understand that there are layers of capabilities that you can build leveraging Generative AI.

As you can assume, majority of companies will start their journey at the productive layer. In tandem with other technologies, Gen AI can help build hyper automation capabilities. These are generally considered to be low-hanging opportunities and are a good way to start your journey. However, as the focus of this report highlights, the need of the hour is to start working on building that apex level group of capabilities.