GenAI helps exploit opportunities in the supply chain

Generative artificial intelligence (GenAI) is going to radically change the way supply chains work. This was evident from Supply Chain Media’s webinar, in which Piet Buyck of Logility explained the impact of GenAI. “With this technology, all relevant information is immediately available to all stakeholders. This allows us to react quickly and adequately to changes, and seize the opportunities that come with them.”

By Marcel te Lindert

Consider generative artificial intelligence (GenAI) as your junior assistant, advised Piet Buyck. “That is someone who can help you well, but does not have the same insight as yourself. A junior assistant can help you prepare for a meeting, gather answers to your questions and turn data into graphs. But a junior assistant is not someone to whom you entrust the final decision,” said the senior vice president at software vendor Logility.

Using GenAI saves labour and may reduce the number of supply chain jobs, but people will still ultimately make the decisions. “GenAI results in higher-quality and content-rich jobs. Yes, we need fewer people because of AI, but they are a lot more productive than before. And that is a welcome development in these times when there is a shortage of knowledge workers,” said Buyck.

iPhone moment

Piet Buyck is co-founder of Garvis, the start-up that a year ago was one of the first supply chain software vendors to deploy GenAI to unlock information from software. Last autumn, this innovation was one of the reasons for Logility’s acquisition of Garvis. Buyck: “Gartner research shows that half of supply chain organizations want to implement GenAI in the next 12 months. We have previously indicated that this is the ‘iPhone moment’ for GenAI. This is the first time that this complex technology is so accessible to businesses.”

During Webinar Wednesday, Buyck explained what GenAI is. “GenAI is the self-learning form of artificial intelligence that provides access to technology through language. This is not to be confused with narrow AI which consists of algorithms for solving specific issues, such as determining stock levels or planning promotions. Yes, we can use GenAI to talk to customers via chatbots and generate software code faster, but the possibilities extend much further.”

Antifragility

The strength of AI is that it can translate the complexity of the outside world into insights that can immediately be shared digitally with all stakeholders. “But we don’t just want to share numbers, but also the story behind those numbers. We want to explain how we get to those numbers, and also to provide a broader context so that we can adapt to events in the outside world,” Buyck stated. “According to research, 89% believe that today’s uncertain and unpredictable world is here to stay. AI can help to navigate this world.”

Buyck mentioned the term ‘antifragility’, which was introduced into supply chain by Gartner. Antifragility is the ability to turn disorder into opportunities, possibilities and growth. “Every change brings risks, but also opportunities. The question is how to exploit those opportunities. To do so, we not only need to make our supply chains resilient and robust, but also to create information that allows us to respond appropriately to change.”

No learning curve anymore

Many companies have made their supply chains more efficient in recent years by centralizing decision-making procedures. But now that supply and demand are not as stable as they used to be, there is a growing need to understand in detail what is happening in every corner of the supply chain. Existing sales & operations planning (S&OP) processes fall short, according to Buyck. “S&OP is an invention of the last century to make the different silos within a company work together. All silos create their own planning, then come together to align it. That leads to a time-consuming process which results in compromises.”

With the deployment of GenAI, that process will look completely different. It is no longer necessary for an employee to dig up data from different systems to create an inventory report. Asking the GenAI application a question in plain language is enough to generate the same insights. “Another advantage is that the technology is in the cloud and therefore scalable. And you can very quickly broaden the scope from, for example, demand planning to other disciplines. Moreover, you no longer have a learning curve like you did when building and constantly refining models. The possibilities are endless.”

Increasing the demand

Buyck specializes in demand planning. To explain the impact of AI on this field, he described demand as the flow of a river. “The demand flows not only to your own company, but also to all your competitors. Many people try to predict as best they can which part of the river flows to their own company. But much more interesting is the question of how to increase the size of that river. For that, it is necessary to understand the determining factors. Or in other words, which events in the outside world, which decisions by customers and which actions by competitors will determine the size of demand?”

To move quickly and seize opportunities, we need to move away from hierarchical decision-making procedures. Buyck quoted the words of former Dutch general Mart de Kruif, who advocates placing responsibility low down in the organization. If you have to make quick decisions on the front line, you don’t have time to wait for the approval of a general based miles away. “You have to explain to people what the goal is, but not how to achieve it. You have to provide the frameworks within which people can decide for themselves how to achieve that goal. AI provides the information and speed needed to make those decisions.”

More sales

Buyck ended the webinar with a real-life example of a drug manufacturer. The company suddenly saw a huge spike in demand for a particular drug because a competitor could not supply it. That information ended up in a spreadsheet that eventually reached the central planning department. The latter wanted to know more about the peak and started investigating, which ended with a phone call to the regional planner asking whether the peak was a one-off or not. “With AI, that same information is readily available to all involved.”

The results of AI are promising so far. Delays in information transfer fall by 70%, while errors in the information decrease by 15 to 30%. The result: sales increase by 4 to 6%. “And lastly, these new systems deliver value much faster. With these self-regulating AI systems, you don’t need dozens of data analysts to set them up.”