Mike Landry, chief executive officer of ketteQ, lays out a path for companies looking to get started with using artificial intelligence in supply chain planning.

There’s no doubt that AI has come to supply chain planning, and is beginning to make its mark. “But right now, we’re just getting started,” Landry says, anticipating a lengthy journey in which the technology will be tested and implemented in all of its “flavors,” including generative and agentic AI.

“There’s going to be a lot of learning, and some missteps perhaps,” he says, “but the next few years are going to be pretty exciting.”

Of particular promise are AI agents, which possess the intelligence to carry out tasks that a human might otherwise have to execute. That won’t necessarily lead to a wholesale replacement of the human planning team, Landry says — think instead of a process wherein a thousand or so automated agents can make recommendations based on huge volumes of data, which are still overseen by human experts.

When linked with agentic AI, the word “autonomous” might seem frightening to human planners who fear the loss of their jobs, but Landry envisions a future in which the machine’s ability to crunch data is combined with insights and context that can only be supplied by people.

In such an environment, the concept of a “center of excellence” might morph into a “distribution of excellence,” wherein decisions are made at the most appropriate place in an extended supply chain.

Adopters of AI must still be on the lookout for AI “hallucinations,” caused by the system’s access to incomplete or incorrect data. But such mistakes should dwindle as the AI model learns the user’s particular business, Landry says, and becomes accustomed to all of the possible scenarios that a supply chain might face.