Stijn-Pieter van Houten, chief product officer for supply chain management planning with SAP, details the state of the art in supply chain planning, which stands to be transformed in the coming years by artificial intelligence. 

A handful of successful use cases is driving increasing interest in the market for the application of AI to supply chain planning, especially as it relates to scenario simulation, van Houten says.

It starts with the ability to run multiple scenarios in the planning application. Agentic AI, even in its early stages, can then infer the next steps that a decision-maker needs to take, based on the conditions at hand.

Eventually, agentic AI will make it possible for the model to make key decisions on planning autonomously, albeit with a human standing by to verify the output. Results are based on a combination of historical examples and business rules programmed into the model that dictate how to deal with a given situation. It’s all correlated to specific configurations that are matched to the types of products that are flowing through the user’s network.

The model gets better and more accurate as it learns, working from a set number of appropriate prompts and prioritizing them according to which ones will led to the best outcome. At that point, “you go into autonomous execution while validating what agentic AI is doing according to business rules,” van Houten says.

He believes the use of agentic AI will bring about a dramatic transformation of the entire supply chain planning process, allowing companies to move seamlessly among strategic, tactical and operational levels, so that they can quickly adjust to unanticipated shifts in the market, and replan accordingly.