Imagine you have a new intern in the office, a bright young man who arrives every morning full of enthusiasm but has never worked in a supply chain function before. While this young man has a lot of promise, it will take a lot of training and mentoring to help him reach his full potential.
Industry experts say such a virtual intern now exists, with one important condition: The worker is not a recent college graduate — or even a human at all — but rather an application of artificial intelligence (AI). Agentic AI is a type of system that can achieve a specific goal with limited supervision using “AI agents,” which are machine learning models that mimic human decision-making to solve problems in real time, according to IBM.
Still in the early stages of development, this technology offers great promise for solving problems in the supply chain sector, which relies on professionals to make repetitive decisions based on large sets of data — think of all the numbers involved in staffing, inventory, routing, deadlines, tariffs, and the like.
An expanding world of agents
In theory, agentic AI can speed up the decision-making process, as computers excel at analyzing data and making quick calculations. But first, the technology faces a critical hurdle: AI needs to be trained in the intricacies of operations such as storage, fulfillment, and transportation. Far from replacing people with machines, agentic AI relies on supply chain professionals to make things clear, experts say.
Sanjeev Syotia, executive vice president and chief technology officer at supply chain software development firm Manhattan Associates, likens the roles played by various participants in the process to those played by members of a symphony orchestra, saying that agentic AI may be able to make decisions and act as a conductor, but still relies on skilled musicians and humans to play the actual instruments.
Earlier this summer, the company launched a suite of AI agents, saying they can take intelligent, autonomous actions to revolutionize and improve supply chain trade execution and user experience. The launch included five specific digital agents: Intelligent Store Manager, Labor Optimization Agent, Wave Inventory Research Agent, Contextual Data Assistant, and Virtual Configuration Advisor. Siotia says there will be more customers in the future.
The company has also launched a tool that allows users to create their own custom AI agents, tools tailored to the unique processes and preferences of their operations. This tool, Manhattan Agent Foundry, is the same platform that the software developer’s research and development (R&D) team uses to build the commercial AI agents the company offers, but is now available for customers to either use themselves or contract with Manhattan or external partners to develop new specialized agents.
According to Sciotia, these AI agents can be used in a variety of common supply chain applications, such as:
- Manage warehouse flows, handling tasks such as balancing inventory reserves, identifying tasks that are running behind schedule, determining whether reassigning warehouse workers will get operations back on track, and then communicating to those workers exactly where they should go.
- Transportation billing tasks, such as assigning AI agents to read PDF files frequently sent via email by small carriers and then entering that information into the shipper’s software platform and taking steps to resolve any issues detected.
- Solve an interpolation problem, such as answering the question “Where are my things?” Questions and tracking of lost incoming shipments.
How to train your new AI assistant
When it comes to getting the most out of AI agents, proper training is key, says Rachit Lohani, chief product and technology officer at supply chain software developer e2open. In Lohani’s view, a common mistake in the market is to confuse automation — which is deterministic and role-based, with zero intelligence — with artificial intelligence, which acts as an assistant or partner.
“Agent AI is about providing you with an assistant. So [to design it]We need to understand your key metrics, how you make decisions, how you rationalize them, and what you are working to improve. “Then when we train the model, we tell it, ‘Start making decisions like this,’ and from there, it keeps learning,” Lohani says.
A hallmark of AI technology is that it improves over time, absorbing lessons about how to improve its performance every time a human teacher corrects it. This means supply chain professionals play a crucial role in providing oversight and feedback to agent AI systems, he says.
“We understand the value of ‘human in the loop,’” says Lohani. “Humans are at the center of this cycle.” “If you don’t have governance and human oversight, you will fail.”
Reserve the right to click “buy”
This is also the prevailing wisdom in the retail sector, where a number of big players are using AI to assist shoppers with their online purchases – for example, by comparing prices, shipping times, and availability; Send them alerts about price drops on items they have viewed or added to their wish lists; Narrow down choices based on historical preferences.
However, there appear to be limits to how much control shoppers are willing to give up. In a report released this summer, retailer Walmart found that although consumers are increasingly open to letting artificial intelligence guide their shopping trips, they are not yet ready to trust a digital assistant to select and purchase products on their behalf..
“There is a strong desire for human-in-the-loop systems where human oversight and control are maintained,” Desiree Goosby, senior vice president of technology and emerging technology strategy at Walmart, said in a statement announcing the study results. Walmart’s Retail Rewired 2025 Report: Agent AI at the Heart of Retail Transformation. “In fact, 46% of respondents said they were somewhat or very unlikely to use a digital assistant or agent to handle their entire shopping journey for them. For me, the preference is clear: AI can help guide, but shoppers want to be the ones who click ‘buy.'”
Walmart concluded that shoppers are looking for ways to “keep shopping fun” while simplifying the practical parts with digital assistants.
Achieving this balance is a clear goal across the industry, whether it’s on websites, in stores, or in warehouses and data centers, technology leaders say. Tasking AI agents with performing menial tasks quickly and accurately could free up humans to do more precise work, jobs that require the kind of creativity, wisdom, and expertise that can’t yet be replicated by code. So, when your digital assistant shows up for his first day at work, remember to be patient, as it will take some time for the new employee to get up to speed quickly.
