Artificial Intelligence continues to dominate supply chain headlines and conversations. Much of the discussion focuses on what AI can do within a company’s four walls, but supply chains are inherently multi-enterprise. How does AI become more valuable when it operates across a Supply Chain Operating Network (SCON) rather than only within a single application or company? That’s the primary question I discussed with Heidi Benko, VP of Product Management and Strategy at Infor, during a recent episode of Talking Logistics.
The Current AI Buzz
Heidi recently returned from a number of industry conferences, so I began our discussion by asking her what she’s hearing about AI. Heidi notes that questions about what AI is and how people are using it are similar to last year. What has changed is that companies are much more focused on generative AI and Agentic AI and how to deploy them.
Heidi says that most companies have moved beyond simply considering AI and now have a mandate to begin leveraging it with their vendors or within their internal processes. So, companies are asking how to get started, which opportunities to prioritize, how to measure ROI, which AI models to use, and how to address issues such as data security, hallucinations, and training. “We’re getting these much sharper, more skilled questions,” she says.
Expanding Value Through a Network
Supply chains are much broader than the processes within the enterprise. They involve many suppliers, partners, and other entities. I asked Heidi how companies can derive greater value by expanding the use of AI across a Supply Chain Operating Network.
Heidi comments that for years companies have talked about breaking down silos between functions and processes across the enterprise. To get real transformation and better serve customers, that same principle must be applied across a SCON. Applying AI on top of that and getting a holistic view of the data across the network allows companies to better handle disruptions and capitalize on opportunities.
AI can analyze network-wide data and help make intelligent decisions. As Heidi explains, connecting all of this information inherently involves high volumes of data and decisions. AI can help identify a disruption and what to do about it, identify an opportunity, quickly synthesize all that data, and connect workflows with the parties involved.
What makes the network especially valuable is not simply that it provides more data, but that it provides broader context. Instead of analyzing a shipment, order, or disruption from the perspective of one company or application, AI can draw on data from multiple systems and trading partners to better understand downstream impacts, available options, and the trade-offs associated with different actions.
Heidi also points to AI agent-to-agent collaboration across different network partners as a potential next step in this evolution.
In fact, when we asked our Indago research community what their biggest challenge was in making faster, smarter decisions, the number one answer was the fragmentation of data across functions, applications, and partners. A SCON is the platform for bringing together that data. AI can gather network data and provide an analytical layer on top.
Creating Value
How can AI create value across a network that would otherwise be difficult or impossible to achieve?
Heidi gives an example of a supply chain disruption, such as a weather event or port congestion. A network can bring together information about the delayed shipment, the orders and items on it, the customers affected, and other relevant data. It can also use network insights to calculate a more accurate ETA and determine the impact of the disruption.
An AI agent can analyze the impact across multiple partners and customers and prioritize what actions should be taken. For example, it can conduct cost trade-off analyses and take actions such as rerouting a shipment or communicating with a carrier. As Heidi puts it, “There are a number of things that it can do, but it really starts with having all that network data and insights versus all those broken-down silos.”
I compared it to using Google Maps or Apple Maps. The application can give you a more accurate ETA and recommend a faster route because it isn’t relying only on your historical travel time; it is analyzing real-time data from thousands of other drivers around you. The same principle applies in supply chain: network-level data gives AI much richer context for evaluating disruptions, options, and potential actions.
AI Leaders and Laggards
I asked Heidi what differentiates those companies gaining value from AI versus those that are just starting out.
She mentions that those who have a focused strategy with a top-down, holistic approach that defines what they want to achieve and the expected outcomes/business value will be most successful. Critically, they look at the decisions that will have to be made, the processes impacted, and how they want to redesign their workflows. Then they assemble a cross-enterprise collaborative team to make it happen. They also define targeted use cases to quickly evaluate results and determine which ones can be scaled.
One of Heidi’s most important recommendations is not to simply automate existing workflows with AI. Companies should use AI as an opportunity to rethink how the work itself gets done. As she explains, companies shouldn’t necessarily try to fit AI into their existing workflows; they should “be open to having AI change those workflows.”
This is an important point because companies have historically fallen into the trap of simply automating broken or inefficient processes. AI provides an opportunity to step back and ask whether there is a better, simpler, or more efficient way to get the work done.
Actions to Take
For companies looking to deploy AI across their networks, what actions should they take to get started? Heidi suggests companies start by creating their strategy, evaluating the decisions to be made, and examining whether they have the underlying data foundation across the network to support their strategy. As Heidi emphasizes, “If you don’t have trusted data that’s accessible in the right way, you’re not going to be successful.”
And when it comes to supply chain, that data foundation shouldn’t end at the company’s four walls. It needs to span the broader network of trading partners, systems, and processes. This is where a Supply Chain Operating Network can play an important role by providing the data foundation for cross-enterprise business processes.
Heidi had many more examples, insights, and suggestions to share during our discussion. Therefore, I recommend you watch the full episode for all of her insights and advice. Then keep the conversation going with your own comments and questions.