Executive Summary:
Discover how AI is revolutionizing supply chain
management in this must-listen episode of The
Enterprise Edge podcast, recorded live at Oracle AI
World 2025. Host Mark Vigoroso sits down with
Derek Gittoes, Oracle's Vice President of Supply
Chain Management with 25 years of industry
experience, to explore the seismic shift happening across global supply chains. Learn how AI agents are compressing multi-week analytical exercises into just 10 minutes, eliminating wasteful manual processes that have long plagued supply chain operations, and enabling companies to respond to disruptions with unprecedented speed. Derek shares real-world examples of autonomous workflows - from interpreting supplier recall notices to analyzing the complete impact of component changes across entire production ecosystems - and reveals why every customer conversation now centers on transformation rather than cost-cutting. Whether you're navigating tariff changes, combating labor shortages, or seeking to future-proof your operations, this episode delivers practical insights on how AI is augmenting human capability and driving the next generation of supply chain resilience. Stream it now, and be sure to LIKE, SHARE, and COMMENT!
​
Transcript:​
Mark 00:00
Hello, everyone. This is Mark Vigoroso, founder and CEO of the enterprise edge. This is the enterprise edge podcast we are recording live at Oracle AI world 2025 at the Venetian in Las Vegas. We are joined today by Derek Gittoes, is vice president of supply chain management at Oracle. We will be diving into the complex, rapidly evolving world of supply chain in the context of Oracle's applications, and everything that, AI means the context of supply chain. So Derek, thanks for being with us. Thank you for having me. Great. So we'll get to get started if I'm if I'm doing my research correctly. I think you started at Oracle around the turn of the century. Yes, around 2000 so it's been century, 25 years. Do you remember? I'm curious if you can think back, what were you working on back then? What was your first couple of projects? And were you with supply chain? Were you focused
Derek 00:57
on supply chain? I've always been in supply chain. Okay, at Oracle, and even prior to that, yeah, and I started off working in the transportation management area. So planning and optimizing shipments and executing transportation activity for got it manufacturers, distributors and others, which is still part of the solution, part of the portfolio.
Mark 01:15
Yeah, absolutely got it in that. With that in mind, how many cloud worlds, open worlds, AI worlds, have you been to and what are some of your observations from this week?
Derek 01:28
I believe I've gone every year for the last 20 years. 20 years. Okay, of course, there was some, you know, covid breaks, of course, in there. But yes, it's an annual it's an annual event. And as far as what we're seeing, or what I've heard in my 30 some customer interactions this this week, you know everyone, if they're not on a supply chain transformation journey, they are planning to start one. And by that I mean looking to modernize not just one part of their supply chain, but literally, you know, modernize their entire supply chain operations, both from a technology perspective, as well as to look at how, how they may actually run their business differently in the age of, you know, new technologies like AI and what that introduces, as you know, possibilities that weren't, weren't possible before,
Mark 02:23
of course, of course. And in that, in that spirit, you know, supply chain is a there's a reason why people can get PhDs in supply chain management, is because it's, it's, it's incredibly complex and, and I'm curious in your travels, and you think about now with autonomous or semi autonomous workflows becoming real in supply chain, whether that's in procurement or supplier assessment or other types of contingency planning. What have you right there's there's more and more possibilities every day for what can be semi automated or fully automated. And I'm curious, where do you think that balance is in a supply chain context between human oversight and AI driven automation, right? I mean, there's, there's a role for both. I think I would say in perpetuity, it's hard to say that, but there's a role for both. But are you thinking about that, or talking about that in your conversations with customers that are going through transformations? Are they thinking about that?
Derek 03:37
Sure, yeah. I mean, I would say absolutely to all of that. So first of all, for as many years, has been supply chain technology and supply chain systems. There is still an incredible amount of activity that happens manually, that if you looked at it from a pure waste or value add perspective, there is no value add, but it's also they've been areas that have been somewhat impervious to automation until now. So for example, classic, a classic example is you receive a document, let's say it's a recall notice for a given product, and now you have to go off and, you know, figure out where you have all that inventory and what you're going to do with it that recall notice comes in different formats, if not, every supplier uses the same format, and so it takes a person and a level of intelligence to interpret that document, because it's not just reading letter by letter, code by code, that it takes interpretation to then Type it into a system large language models being implemented in an AI agent can do that same interpretive work. It's not just reading word, this word and that word. It's, as we say, a semantic understanding of what's in that PDF document or in that email to then drive the rest of the transaction processing. So I would say that's an example of, you know, using AI to automate. I mean, generally speaking, you don't want to have your highly paid staff doing a bunch of email swivel chair integration. You would rather have them figuring out, you know, how to solve a customer service issue, or how to improve the quality of your products, or how to, you know, launch new products into into the market. So there's just a credible amount of waste that can still be taken out. And there's still, you know, there's opportunities to look at doing processes in new ways, ways that you couldn't do before, by providing sort of a intelligence and connectivity between, let's say, what you're hearing from your customers and the feedback in your product versus what your product designers and engineers are being able to oh, I'm now getting the feedback, you know, how my product is doing in the field, and therefore I can make, you know, adjustments to, you Know, design and deliver even better products.
Mark 05:59
So in you know, I like to think about supply chain as is like big multi variable math equation, right? Where you're trying to optimize across sometimes conflicting objectives, right? You're trying to figure out cost drivers, quality, risk, exposure, resilience and a million other things, compliance, right? There's a lot of different variables, and I'm curious how Oracle in the supply chain, software development area is thinking about orchestrating potentially multiple agents that are that should be sort of harmonizing, right? They should, they should be doing that math faster than any human could possibly do, and optimizing faster than any human, or perhaps previous software applications could do. So is that, as that an oversimplification, or how are you thinking about the sort of the multi variable optimization and supply chain in the AI era, with multiple agents that are coexisting.
Derek 07:06
Yeah, I'll give you an example of one of the AI agent teams, because and in the in the world of agentic AI, you can define agents who have a particular task or role objective, and you can have a team of agents that are working literally, collectively on a bigger process. So one example is, you know, I get a notification from, you know, a supplier of one of the components I use in the products that I make. And notification is, this component is going end of life, and we're going to replace it with this new component. All right, now I want to be able to determine what is the full supply chain impact of now that I have to work this new component into my production, how many bills of material is that going to impact? How much of that supplier's existing product do we have on hand, and where is it, and what should we do with it? Should we use it all up, or should we, you know, return it? How much work in process do I have in manufacturing that's already underway? How much current booked or forecast customer demand do I have for that we work with customers on that particular use case where for one component change, it can literally be a multi day, multi week analytical exercise, versus a team of agents that could do it in 10 minutes. Yeah, it's incredible. So the potential, and it's not that people don't know how to do the math. It's just when you have data spread out across many systems, and you have to sometimes physically go around and talk to people. And these people need to run their reports so they can give you their number, and you take that number and you put it into this thing, agents are mimicking that. It's just they, they can do it fast, far, far faster, yeah, and they'll keep on, they'll be doing it, you know, Friday night, after everyone's gone home, and they'll, you know, be doing it, you know, over the weekend when no one's there. So it's a, it's a you can produce, not only, I would say, maybe not necessarily, better decisions, but you can make more consistent decisions, and you can make them quicker, yeah, quicker, as in, like, orders of magnitude quicker, not a few percentages
Mark 09:27
definition, yeah. So, you know, there's a lot of discussion about vertical versus horizontal, right? So when you think about the various industries that have global, complex supply chains. You know, there's quite a lot of variability across chemicals versus automotive versus aerospace versus, you know, lots of different categories, right? And I'm curious, how does Oracle approach vertical uniquenesses and idiosyncrasies that are must haves. Maybe some of them are geographic driven. Maybe some of them are regulatory, geopolitical security driven, right? There's different levels of sensitivity around personal data and all that stuff that's unique to industries. So when, when companies that are out there that claim to be completely verticalized, like there are ERP providers that are only for mid market manufacturers, period, and there are only for apparel companies, period. Oracle, obviously, you could say definitionally is more of a horizontal but how do you address the vertical needs that you see
Derek 10:38
out there? Yeah, it's great question. So, so first of all, there are definitely business requirements that vary by industry. Vertical. Just as you said, if I'm an aerospace manufacturer, that's not the same as if I'm, you know, a medical device manufacturer or chemical or what have you, our strategy is, however, we have horizontal products. So I'll take manufacturing as example. We only have one manufacturing component of the fusion supply chain suite, but we have built in, through our experience and years that we've been at it, the flexibility to support what we call different modes of manufacturing. So if I'm a discrete manufacturer making one at a time, or I'm a process manufacturer, where it's a continuous production process, we can model that. We can abstract those characteristics away so that it doesn't matter whether you're it's a discrete process or a process or a discrete manufacturing model, or a process based, in fact, in some cases, the companies will be both, yeah, it's a process based, until you get to the end where you're doing packaging, and ultimately it becomes a
11:53
discrete
Derek 11:54
item, a discrete container of fluid, yeah? So, and we've, taken that approach across all that we do, whether it's in manufacturing or order fulfillment or planning, and again, focus on needs of particular industry verticals. And how do we accommodate that through configuration of our software, as opposed to having six different versions of a manufacturing product or six different versions of a logistics product.
Mark 12:26
I mean, somebody's probably done the study that says, sort of, if you boil it down to like the atomic level, are supply chains really more the same than they are different across industries?
Derek 12:39
Well, I would say customers may not want to hear it, no, but the answer is, supply supply chain functions are incredibly horizontal. Yeah. I mean, not to take away the fact that there is, you know, variability, but I'll take so for example, all supply chains involve products by definition, so you need to have a product record foundation, foundation and how you model products and the attributes of the product. Those attributes cover, if you're a manufacturer, those attributes cover how you make the product, the components that go into the product. Obviously, everyone sells their product, so how do you model the attributes around? Well, what are the valid configuration options for this product, or what's our pricing for that product. So everyone has products. It may look very different. One product might be a data center, and the other product might be, you know, a pair of shoes. But at the end of day, it is, it is product. So that's how we're able to, again, horizontalize the solution, but provide again. When I'm again, our product model may have, like, hundreds and hundreds of product attributes, but for a given company in a given industry, maybe only 10% of them are actually relevant to them, and then that's the that's the percent. That's what they use in the solution set.
Mark 13:57
Got it. It's interesting. I'm curious the comment that you made when we started, which is, most companies you're talking to are going through or about to go through supply chain transformation of some kind, which probably includes more than just technology, right? It probably includes processes and org structure and skills and talent, and, you know, the whole classic people process, technology and data, right? The All, everything. So on the people side, when you think about the world in which humans and agents coexist, are you hearing from customers that they are actively thinking about rebalancing human employees, upskilling, redeploying, a lot of that wasted, busy work that could now be picked up by agents. It's probably going to free up some capacity, right? So is that, I know a lot of people fear that that automatically gets taken to the bottom line in the form of workforce reduction, but I'm hearing more that it's it's not it's being there. They're redeploying and they're upskilling.
Derek 15:03
And I have not had a single conversation at this event or one that I can think of, yeah, where the motivation for the transformation of our AI agents was to get rid of people? Yeah, exactly. Because, literally, one of the number one problems in supply chains is a demographic problem. You do not have enough people, there's a labor shortage to do the work. So the last thing you want to do is get rid of the people who you have. Instead, you want to use, whether it's AI agents or other technologies, to make your people more productive, either by allowing them to do do more of what they're doing in the same or less time, or to have them work on other things that they otherwise don't have time to work on because they're, you know, typing in reading emails and typing into systems. So this is about reaching significant improvements in productivity and quality so they can grow their business, ship more product, make more product. It's very much with a growth being approached with a growth mindset, as opposed to, oh, this is a way we're going to cut costs.
Mark 16:13
Yeah. It's more like an irrational fear. I hear it from, you know, in business settings, but also sort of, you know, propagated in consumer media, where there's just people, where the Rise of the Machines, they're going to take over, right? I liked what the gentleman from Avis, I think it was Avis Budget. He redefined AI as augmenting individuals. Yes, I love that. I thought that was awesome. And I've stolen it like five times since then. But it's, it's a, I think it's a way more accurate description of what the technology is actually doing. Yes, it's not really art, and I guess it is artificial, but it's augmenting your existing workforce. Is what it's
Derek 16:51
doing. So it is a force multiplier, what you can get out of people and again, either by taking away again, who wants to read emails and type them in, yeah. I mean, no one goes to Yeah, No one wakes up with that. That's their objective, right? Objective in life and and certainly for, you know, again, newer generation, you know, current generation, you know, I'll take my daughters, for example, in college or graduate school. You know, they're very comfortable using the AI assistants and chatbots to do their to do their work, develop their code, to help write their papers. Yeah, I mean, to help write, not to write the paper, to assist in the proofreading, copy, Ed, copyright, yeah, it's a thought like a thought partner. Yeah? These become literally, again, you think of them as assistants to help you do the task, as opposed to a threat that's going to remove me from the equation.
Mark 17:50
So two couple more, I'm going to steal the idea that, well, in my experience, companies that are truly global, multiple geographies, 1000s of people. They're not a homogeneous technology stack. There's lots of different systems. Some of them are legacy systems. Some of them are industry specific, off the shelf, homegrown. It can be a mess, right? And so I'm curious, how is Oracle accounting for that in terms of some of the integration challenges that exist in supply chains, where you're not only dealing with your own enterprise, but you're dealing with suppliers, multiple tiers of suppliers, and multiple geographies, systems integration challenges that are incredibly complex. Are you? AI is not making this an issue. It's always been an issue. But I'm curious, how does Oracle think about this problem in supply chain, and integrating and offering a total solution, even if you're not on homogeneous Oracle
Derek 18:51
shop, yeah, so I mean, in the ideas, ideal solution for our global, vast, distributed companies is that they do get to a single, literally single instance of our Fusion Applications to run all those different functions and make life the easiest. Of course, the reality is they may not be able to do that, or they might not be able to do that right away. So we've built in interoperability, even though we have an integrated suite, yes, any of those components that we have can be replaced with some other software component. Got it. So if you just want to use our logistics part of the suite, if you want to keep your legacy manufacturing, you can do so. If you want to use our product lifecycle capabilities, but keep your legacy order management, fulfillment, you can as a starting point. Or maybe you put fusion modernize everywhere except for this one country, because, you know, they, they run things so differently. We just don't want to touch what we're doing in that in that part of the world.
Mark 19:56
Got it Okay. Last, last question, and we'll let you go. Quick question. So, resilience, supply chain, resilience, became somewhat in vogue after covid, right? And I'm curious in the years since then, what, what has what has come to production for Oracle or been released to specifically advance that cause of resilience? Are there any, like maybe one or two examples of tangible capabilities that are contributing to that objective of being more resilient and responsive and not so reactive.
Derek 20:29
Yeah, so I would say it comes in a couple of areas. One is through planning capabilities, so being able proactive to understand, for example, there's a tariff change, how can I model the impact of that tariff change, either in my demand or availability of my supply? So being able to again, proactively model impacts before the impacts arrive, and the second is through the introduction of AI agent technology, so again, able to execute the analysis, understand the impact, and then actually, you know, respond to that in a much faster way. Got it well.
Mark 21:05
Derek, thank you. I've I've left you two minutes to get to your next meeting. Thank you so much for your time and your your candor. I appreciate your experience and your insights. Thanks everyone for listening. Please be sure to share and like and comment. We value your input. Until next time this has been the enterprise edge podcast, and we'll catch you for one more before the end of Oracle. Ai, will AI world Take care, everybody. Bye. Bye.​
