As supply chains grow more complex and disruptions become increasingly frequent, warehouses are under pressure to move beyond reactive operations toward intelligent, real-time execution. AI is emerging as a key enabler, helping organizations automate routine tasks, improve decision-making, and respond faster to operational challenges without disrupting existing workflows. In an exclusive conversation with AI Reporter America, InfiosRichard Stewart, EVP Product & Industry Strategy, Infios shared how its latest AI-powered warehouse innovations are transforming execution through practical AI agents, enabling businesses to streamline operations, enhance workforce productivity, and build a more adaptive and resilient supply chain.

1. What warehouse challenges drove these latest AI-powered innovations?

Traditional warehouses are built to be reactive. They are systems of record that are great for telling you what happened yesterday or what we think is happening now but that’s not enough in today’s reality. Today, warehouses deal with complexity that never could have been imagined when these old systems were put in place: labor constraints, inflation, regulatory demands, rising consumer expectations and countless other constantly evolving issues.

Warehouse teams are already operating at full capacity before disruptions hit. Managers are juggling countless tasks, and they don’t have the time to commit to manually solving all these problems. Constantly reacting to fires after they happen keeps the manager at an executional level rather than strategic and that prevents the warehouse from working at its highest capacity.

The other major issue with the traditional warehouse model is siloing of information. For example, if a truck is coming late, maybe the supervisor knows he needs to adjust the labor plan but he hasn’t had time to tell the employees before he gets pulled into something else. Too much information ends up living in someone’s head without being passed along to the rest of the team, cutting down efficiency and slowing down processes.

That context is what gave us our focus when creating our AI agents. We want to be able to provide practical, useful innovations that provide a clear use case, whether that be in time savings or reducing silos to increase the flow of information in order to make decisions and act more quickly. These AI capabilities are meant to reduce manual coordination and friction in daily warehouse operations, leading to provable better results today.

2. How do these capabilities improve decision-making and operational efficiency on the warehouse floor?

The main efficiency driver with these AI agents is speed. The processes that used to take hours or even days can now be done in minutes. For example, look at our knowledge assistant. Without this assistant, you would be forced to read through documents, skimming and searching for the information you need. But with our AI agent, it becomes a simple natural-language search and response in seconds. This not only saves time, but it improves consistency and adherence to policies. The easier it is to access knowledge—and let’s face it, some of these documents can be pretty dense—the more likely it is to be used and followed.

By cutting time needed for some of these previously manual tasks, it creates more time for strategic thinking to move the business forward. Infios AI enables these warehouse leaders to proactively make decisions focused on how to create efficiencies and optimize the warehouse.

But saving time on routine tasks is only half the equation – what truly separates high-performing warehouses is how quickly they respond when things so wrong. With our error resolution assistant, operators are able to quickly and accurately diagnose and resolve issues. Since they receive a root cause analysis, the resolution can be found in a fraction of the time, getting the warehouse back on track. Less time spent fixing errors means there is more time for value-added work to be completed.

At the end of the day, the goal of our AI agents is to take some of the work off the plate of the warehouse managers to free them up to do whatever is needed. We know that no two days are the same in any warehouse—with Infios AI, we are able to automate the everyday tasks and solve the disruptions to create a smoother operation.

3. What differentiates Infios’ AI approach from traditional warehouse automation solutions?

I previously mentioned this but it’s worth saying again: the biggest differentiator between traditional automation and Infios AI is speed and efficiency. By embedding practical agents directly into the workflows, an Infios AI-powered warehouse is able to combine the knowledge base with real-time data and intelligent automation to create a more agile and streamlined warehouse operation.

One thing we are really trying to do with all the AI agents—both live and those on our roadmap—is making sure they have a practical application. It’s easy to get caught up in the number of agents you have or how many bells and whistles they have but at the end of the day, adoption of these tools is going to be based on practical, sensible uses. We are striving to create AI that solves the problems we hear from customers. By focusing on the root problem—siloed systems, time-consuming tasks taking away from big picture thinking—we are creating clear ways these agents will make your day easier if you’re a warehouse operator.

We’re also trying to build agents that work with the current human processes instead of trying to reinvent them. Error resolution, product documentation searches and labor coaching are all tasks that are currently happening with processes in place. Infios AI does not reinvent these processes, we instead make them more efficient. This is part of our focus on execution without interruption.

4. How do AI agents help bridge the gap between visibility and execution in supply chain operations?

Most organizations already have visibility tools like dashboards, alerts, and reporting solutions that tell them what's happening. The gap is in what happens next, and that's where the majority of supply chains are still failing.

When an inbound truck is running behind, a system can flag it. But translating that flag into a labor replan in the warehouse, a customer notification, and a re-tender decision for different carriers, across three separate systems, still requires humans to manually connect the pieces. The alert exists. The action doesn't. That's the visibility-to-execution gap, and it's where disruptions cascade into margin erosion and missed service commitments.

To close that gap, AI agents need to operate inside execution systems, not above them. They don't just observe; they act. When Infios AI detects an inbound delay, it doesn't surface a report for someone to act on tomorrow. It assesses downstream impact across warehouse, order, and transportation systems simultaneously and initiates a coordinated response—within guardrails the customer has defined.

This is why the agent architecture matters. In a warehouse context, our inventory error resolution assistant doesn't just identify a missing item—it traces root cause, identifies the pending replenishment task, and recommends the recovery action. The operator reviews and approves. The human stays in control, but the investigative work that previously took hours happens in minutes.

The same principle applies across the execution chain. Labor coaching agents surface performance gaps and deliver individualized guidance during the shift, not after it. Knowledge assistants surface direction in seconds rather than requiring someone to ask their supervisor or leave the floor to search through manuals.

Execution is where supply chains win or lose. Visibility is the input. Coordinated action—across systems, in real time, within defined policies—is the outcome. AI agents, run and managed by Infios Archer, are the mechanism that finally connects the two.

5. How do you see AI reshaping warehouse and supply chain execution in the coming years?

The shift that's underway isn't about adding AI features to existing systems. It's about changing the fundamental operating model of execution. For decades, warehouse and supply chain systems have been built around a planning-led execution model. They also relied on a records-first paradigm: plan warehouse inventory, labor and fulfillment needs, capture what happened, present it to a human, wait for direction. That model worked when the pace of change was manageable and plans could change fast enough to keep up with changing conditions. It doesn't work now. Disruptions arrive faster than manual processes can absorb them, and the labor pool required to manage exception volume at scale simply doesn't exist.

What we'll see over the next several years is execution becoming a continuous, adaptive loop—where systems sense changes, coordinate decisions across domains, and take action without waiting for human intervention at every step. That doesn't mean removing humans from the picture. It means redefining where human judgment is most valuable: on the decisions that require context and experience, not on the operational coordination that should happen automatically.

In the near term, adoption will center on discrete, high-value use cases—labor coaching, inventory error resolution, documentation search—where the ROI is clear, the risk is contained, and the change management lift is manageable. Organizations will build confidence in AI-driven execution incrementally, expanding autonomy as outcomes validate the approach.

Over time, the organizations that have invested in cross-domain intelligence—connecting order, warehouse, and transportation into a unified execution layer—will operate at a fundamentally different level. Their systems will anticipate constraints before they become disruptions. Their teams will spend less time firefighting and more time on the decisions that drive real competitive advantage.

6. What challenges do organizations face when adopting AI-driven warehouse workflows at scale?

The barriers to AI adoption in warehouse operations are more structural than technical. Most organizations understand the value proposition. The obstacles are what stand between that understanding and production deployment at scale.

The first, and often underestimated, is the need for an overarching solution that can run, coordinate, and govern agents across the warehouse operation. Individual agents can address specific workflow needs and deliver clear value quickly. But without a foundation that manages how those agents operate, that shares context across workflows, controls approvals, and provides visibility into decisions and actions, organizations hit a ceiling. Agents proliferate without coherence. Automation expands without governance. What starts as a focused pilot becomes an ungoverned sprawl. A shared intelligence layer like Infios Archer is what scales agents safely, governs them and connects them into more workflows over time. Without it, agents remain productive, but operate in isolation.

The second challenge is data readiness and usable operational context. Enterprise AI implementations have historically required months of data normalization before a single model can run, building data lakes, mapping schemas, reconciling how the same concept is defined differently across OMS, WMS, and TMS. Especially for a midmarket operator running a supply chain team of eight, that timeline is a disqualifier, not a delay. Architectures like Infios AI fundamentally change that equation, because they can leverage existing data in warehouse, transportation and order management systems to build shared operational context dynamically, without requiring long, costly data engineering projects.

The third is trust. Operations leaders are accountable for service levels and cost performance. Handing decision authority to a system they don't fully understand creates real organizational risk, particularly in high-stakes scenarios like labor deployment or carrier selection. Adoption accelerates when AI operates within explicitly defined guardrails, every action is explainable, and teams can expand autonomy incrementally as outcomes build confidence. Graduated autonomy isn't just a product feature, it's the adoption model.

Organizations that solve for all three challenges (platform foundation, data accessibility, and governed autonomy) are the ones that move from pilot to production and keep expanding towards coordinated execution from there.

7. How do these innovations support Infios’ long-term vision for Intelligent Supply Chain Execution?

The broader market has invested heavily in planning and visibility, but less in execution. Those capabilities have genuine value, but they don't resolve a missed carrier pickup, a labor shortage mid-shift, or an inventory discrepancy blocking a high-priority order. Execution is where the plan meets reality. When those don't align—and they never align perfectly—the question is how quickly and accurately warehouse operations can respond.

Intelligent Supply Chain Execution means transforming systems of record, such as Warehouse Management, into a system of action. That requires three things: a shared intelligence layer that understands data across OMS, WMS, and TMS in real time; agents that act across those domains simultaneously rather than in sequence; and a trust framework that governs how autonomy expands as outcomes validate the approach.

The warehouse innovations we've released—labor coaching, inventory error resolution, documentation assistance—are purpose-built expressions of that architecture. They're not standalone features. Each one represents the Infios Archer intelligence layer operating inside execution, at the specific decision points where warehouse teams are losing time and accuracy today.

The long-term vision is execution that is continuous, coordinated, and adaptive—where a disruption in transportation triggers an automatic labor replan in the warehouse and a delivery promise adjustment in the order system before anyone has to intervene manually. Where every decision is backed by cross-domain context, governed by policy, and fully explainable after the fact.

What we're delivering today builds the foundation for it—use case by use case, agent by agent, as organizations grow into greater autonomy at their own pace.