For those of us who have been following Blue Yonder's transformation since Panasonic's acquisition in 2021, this moment feels significant. After watching CEO Duncan Angove methodically rebuild the company's technology stack from the ground up over three years, his long-promised vision of an end-to-end, interoperable supply chain platform is finally materializing at ICON 2025 in Nashville this week. 

The timing couldn't be more fortuitous. As companies worldwide grapple with supply chain vulnerabilities exposed by COVID-19 and continuing geopolitical tensions, Blue Yonder's substantial investments – now totaling $2 billion, double what was initially announced – have positioned the company to capitalize on the emerging agentic AI wave and help enterprise buyers deal with ongoing uncertainty. Speaking with diginomica ahead of the event this week, CEO Duncan Angove said: 

It was two years ago in Las Vegas, my first ICON, and we announced this innovation shock wave, where I said we were going to spend a billion dollars. 

“It was all about putting the product right. It takes time to hire engineers and to ship, but I think the release we had in December was the first really momentous thing where all the pieces came together. And then our release in March, and then the one this summer, it's all sort of come to a head. It's almost like we've got too much to talk about.

What makes this moment particularly significant is how Blue Yonder has patiently built the essential scaffolding – the Snowflake data cloud, a common platform, interoperable applications, and now a knowledge graph – while many competitors rushed to slap generative AI capabilities onto legacy architectures. For those that have followed Angove’s career, this was similarly a successful playbook that he followed at Infor (marketing spin was swapped for doing the hard work of product redevelopment in the cloud). The result is that while others have talked about AI agents for a year already, Blue Yonder quietly constructed the foundation that could make them useful in the complex, high-stakes world of supply chain management.

And if the last three years have been about getting the building blocks in place, this next phase will be markedly different.  With the infrastructure now established, we're seeing Blue Yonder pivot from construction to execution – from promising potential to delivering measurable value through five domain-specific AI agents designed to enhance operations across inventory, retail shelving, logistics, warehousing, and multi-enterprise networks.

The long road to delivery

When I sat down with Angove in London last October, he described a massive change program underway at Blue Yonder, backed by significant capital from Panasonic. The CEO wasn't shy about acknowledging the challenge he'd set for himself and his team. At the time he said: 

Our central thesis was: all of this should be joined up as a set of applications that are designed to work together - that fundamentally enables better cross-functional decisioning, collaboration and orchestration. My worry was, oh my gosh, this is an insane change management exercise at an industry level. An industry that's bought and implemented software like this for 30 years.

Now, that exercise is bearing fruit, with the underlying technology stack coming together and customers beginning to adopt the platform approach in meaningful numbers. He added: 

All the building blocks, getting Snowflake done, the platform built, the first wave of cognitive, actually having all the products integrated so they work together. And we're seeing it show up in our business now. Last quarter, in Q1, we did 135 multi-product deals. In the old days this was a point solution market - you'd just be WMS, you'd just buy planning. And so for the first time ever, I actually saw, wow, the market's interested in buying a set of applications, a suite on a common platform. 

A deliberate approach to AI

As already noted, many enterprise software vendors rushed to announce generative AI capabilities in 2023 and 2024. Blue Yonder, however, has taken a more measured approach. Angove recognizes that supply chain management demands precision that generative AI alone couldn't immediately provide. He explained: 

So when this came along, it was a much easier use case for other categories of enterprise software, because they're not deterministic and as precise as you are in supply chain. So we weren't rushing to suddenly say, 'Hey, we're the supply chain agent platform for the world' like everyone else did. We were very pragmatic about it, and we worked with customers in the field to make sure that there was genuine value. And we didn't just want to reinvent what the numbers based ML was already really, really good at. And that took about a year, and there were a lot of learnings from applying it. 

This patient, customer-focused approach reflects a reality that many AI enthusiasts overlook – in critical operational environments like supply chains, mistakes can be extremely costly. As Angove bluntly puts it: 

People are spinning it way ahead of where it is. It still needs a lot of hand holding and tuning and refinement and all of that. The idea that you can just let this thing go…particularly the autonomous stuff and the agent-to-agent stuff - and by the way the narrower you pick a domain and a certain problem, the easier you are to solve it…so we're just very cautious in how we're scaling it, because mistakes in the supply chain are very expensive.

Welcome realism in an industry that typically favors hype over reality. 

The result of this cautious approach is what Blue Yonder is now unveiling at ICON 2025 – domain-specialized AI agents built on a foundation of its existing machine learning capabilities. Angove said: 

And we're at the point now where we're announcing our agents. We've continued to build out the platform for agentic, which basically is our Blue Yonder orchestrator. We connected that to all of our existing solvers and optimizers in ML. We connected it to all the APIs, so it can see what's going on in a warehouse, in transportation, in inventory - and it's bi-directional, it can go back and execute. 

And then we built fine tuned language models, large language models, into each of these domain cases. They were trained on what we do, so tons of RAG basically went into it, and that's where a lot of the work was. So these are proprietary, fine tuned, highly specialized, large language models for a particular domain in supply chain.

AI agents and knowledge graph

At ICON 2025 in Nashville this week, Blue Yonder unveiled its long-anticipated AI agent strategy, showcasing five domain-specific agents designed to tackle targeted supply chain challenges. As noted already, the company's methodical approach to agent development reflects Angove's careful strategy, avoiding the rush to market that characterized many competitors' AI announcements over the past year. The five specialized agents introduced at the conference include the Inventory Ops Agent, Shelf Ops Agent, Logistics Ops Agent, Warehouse Ops Agent, and Network Ops Agent—each addressing specific operational challenges across the supply chain ecosystem.

The Inventory Ops Agent helps planners match supply with demand by identifying mismatches and exceptions, diagnosing root causes of supply issues, and recommending actionable solutions like alternate sourcing or expediting. For retail environments, the Shelf Ops Agent enables planners to rapidly perform planogram edits with simple natural-language interactions, boosting productivity in what has traditionally been a time-consuming process.

The Logistics Ops Agent addresses the complexity of transportation management by monitoring conditions, recommending route changes to prevent delivery disruptions, automating appointment scheduling changes, and identifying ways to optimize transport costs, on-time deliveries, and emissions.

Meanwhile, the Warehouse Ops Agent helps warehouse leaders by intelligently coordinating interdependent tasks for faster, more confident decision-making. It dynamically reallocates labor based on shifting priorities, optimizes warehouse layouts using predictive insights, identifies outbound risks early, and streamlines trailer operations according to content urgency and staffing levels.

Completing the agent suite, the Network Ops Agent monitors operations across the multi-enterprise network, proactively manages disruptions, and optimizes inbound supply and logistics through automation of order confirmations, stockout resolutions, and carrier assignments, allowing businesses to pivot quickly to avoid delays.

To further strengthen its AI capabilities, Blue Yonder announced a partnership with Microsoft's Azure AI Foundry, a platform for designing, customizing, and managing AI applications and agents. The company will leverage Azure AI Foundry to continue building and refining its agent ecosystem.

Perhaps most intriguing is Blue Yonder's collaboration with Snowflake and RelationalAI to create a supply chain knowledge graph that enhances its Cognitive Solutions. This knowledge graph records business relationships and processes in human-readable form, enabling users to engage with complex supply chain data more intuitively while empowering AI agents to reason more effectively.

The warehouse ops breakthrough

Among the five agents being launched, Angove seems particularly excited about the warehouse operations agent, which he developed in partnership with PepsiCo. He said: 

The warehouse ops agent, it's unbelievable. A warehouse is constant firefighting, right? You're managing labor, inventory orders, trucks coming in, trucks going out - it can see all of that chaos and make recommendations for you. It allows that warehouse manager to rise above all the firefighting. It handles all of that chaos for you. It's really, really cool, because it's available on a mobile phone, but also CarPlay. 

Here is the vision: a warehouse manager is driving to work in the morning. It's 8am. And as you know, in retail, the orders are pushed down about 2am. And as you're going into work, the warehouse ops agent is briefing you. Here's what happened last night. Here's what went out. Here are the picking issues, here's the labor. Here are the things we have to watch out for today. Here are the things I recommend. And literally, by the time you've got to the building, you've had a full brief, you understand what's happening today. You know the challenges, and you know the recommendations you're going to go with.

The early feedback from potential customers, according to Angove, highlights the pent-up demand for these solutions. He said that he pitched the solution to one of the largest 3PLs in the world, where he had the three CEOs across the three regions on. And 30 seconds in, the North American CEO’s response was clear: 

He said: Just 'Stop. I want it, I want it now, I don't care if it's buggy and crashes, give it to me now'. 

Breaking down silos through agent collaboration

What makes Blue Yonder's offering interesting is that it has done the work to enable agents that can collaborate with each other. Many of the solutions we see in the market at the moment focus on singular, uncomplicated workflows - and wouldn’t be equipped to deal with a supply chain environment. However, thanks to the one data approach on Snowflake and by creating an interoperable suite of applications, Blue Yonder has the opportunity to show how agents can work together to solve problems. Rather than creating new silos of intelligence, the company has designed its agents to work together, mirroring the benefits of human teamwork. Angove said: 

What we didn't want to do is, and I've talked for years about the silos you want to break down, so the last thing you wanted to do was build an agent that's siloed. So these agents are actually able to collaborate with each other. And we're going to show one where our network ops agent, which sits on top of the Blue Yonder Network. It can see an issue, and it can actually work with the inventory ops agent and the logistics ops agent to basically autonomously resolve it.

Blue Yonder is incorporating emerging standards for agent communication into its platform and Angove points to how agents working together and working with humans can deliver value beyond traditional teamwork: 

We're supporting Google's new A2A protocol, which is the agent-to-agent protocol. So we're going to announce support for that. There was a really interesting study that came out a few weeks ago that talked about how agents actually deliver the benefits of human collaboration and teamwork. 

And the whole idea here is when you have an individual from a domain and you bring people from other domains into it, it's like design thinking, the quality of the solution is better, and it's informed by people with different skills and domains. So you can imagine if you're a warehouse operator, you're a human and you have other agents, logistics agents and inventory ops agent that are helping you decide alongside you, it actually delivers the benefits of human collaboration, but agentically.

Real-world impact: minutes not days

To illustrate how Blue Yonder’s integrated approach translates to business benefits, Angove shared a compelling example from the automotive industry, which has traditionally led supply chain innovation. He said that the vendor won a deal with a large automotive OEM, one of the world’s biggest, which wanted to deliver its supply chain end-to-end, because it realised that its previous environment, which had been cobbled together, wasn’t going to work. 

He described a complex scenario showing the platform in action:

So this is how it works. It's got Flexis in it. Flexis is doing the sequencing and slotting on the production line. One Network is connected to it. And, for example, a supplier that provides heated seats communicates that it's going to be a few days late. Flexis is connected to the network, which automatically reslots and re-sequences the production line to take heated seats out of it. Flexis communicates to our WMS, which is also on the front end of it, and it takes it changes the components and the work orders that it would have done, because heated seats aren't there in places with non heated seats. 

TMS is updated on the back-end because that's also part of this, to actually reschedule what dealers you're going to go to, because the cars going there are completely different. And sitting above all of it, they've subscribed to our entire cognitive planning stuff, which re-plans all of this. By the way, that would normally take three to four days. That happens in minutes now.

This reduction in response time – from days to minutes – illustrates the real value proposition of Blue Yonder's integrated approach. In an era of constant disruption, from COVID-19 to geopolitical tensions to potential new tariffs under the incoming Trump administration, this kind of agility will likely not only be welcomed by buyers, but sought out by those recognizing that predictability and consistency in the global economic environment are things of the past. 

And with the technology foundations now largely in place, Angove identifies adoption as the next critical phase for Blue Yonder:

Now it's about adoption. We've got cognitive solutions. We have the platform. And like I said, I was really pleased with what we saw in Q1 and multi-product deals. People are buying into it.

The company is already seeing significant progress in accelerating implementations:

One of the things I was really pleased with last quarter is, three years ago when I spoke to you, we were doing two go-lives a day. We're now doing six. It's triple.

The market appears to be validating Blue Yonder's approach, with the company reporting strong financial performance in its recent earnings release. FY24 company revenue reached $1.36 billion, including 14.2% SaaS revenue growth year-over-year.

My take

While Blue Yonder's transformation has required significant investment and patience, the timing couldn't be more fortuitous. The global disruptions of recent years have exposed the weaknesses in traditional supply chain approaches and accelerated demand for more resilient, integrated solutions. As Angove said: 

If you're lucky, you capture computing transitions that you get to ride. And this one is definitely AI, the rise of the data cloud as a next generation stack, and a category of supply chain that's never been platformed and brought together.

The trade policy uncertainties surrounding the recent US election and incoming Trump administration may create short-term challenges but likely strengthen the long-term case for Blue Yonder's approach. As Angove noted during our conversation, In the long term, this uncertainty and unpredictability will be good for a vendor that aims to help buyers navigate against that. How much of this trade war becomes permanent, we don't know. But the reality is it's made companies realize that they need a fundamental restructuring of their cost model in supply chain. 

Increasingly volatile market conditions create a compelling argument for the kind of end-to-end visibility and speed of response capabilities that Blue Yonder is offering. As Angove noted: 

The reality is, the way to combat uncertainty is just being faster and more precise and being technology enabled, so that you can be more agile and adapt. All of that lends itself to more of an end-to-end kind of solution. Longer term, these are actually good things for us.

What's clear is that the foundations Blue Yonder has put in place over the past three years – the Snowflake data cloud integration, interoperable applications, and knowledge graph – position the company well for the agentic AI era, regardless of how rapidly or gradually that transition occurs.

For enterprise technology buyers and supply chain managers weighing their options in what remains a crowded market, Blue Yonder's deliberate, foundation-first approach offers a compelling alternative to the hype-driven AI strategies prevalent elsewhere. The next year will reveal whether that patience and precision engineering delivers the performance that complex global supply chains demand.