Name: Closing Keynote: Driving Supply Chain Excellence and AI at Scale
Uploaded: 2025-03-25T19:11:04.603Z
Duration: 20 min 12 s
Description: Closing Keynote: Driving Supply Chain Excellence and AI at Scale

Transcript for "Closing Keynote: Driving Supply Chain Excellence and AI at Scale": Hi, Suja. Hi, Rohini. How are you doing? Yeah. Thanks for much. Thanks for making it out. Yes. We've had a few technical challenges, but thank you for bearing with it. And, we're really pleased to be here at the closing keynote with, Suja and Sekaran. Suja, just a quick introduction, is a global business operator. She's a P and L leader and digital transformer with deep roots in supply chain health care. Suja serves as a board member on the boards of public companies such as American Eagle Outfitters, AEO, and Cardinal Health, as well as on private companies like Pando, dot ai. As an operator, she served as senior EVP, chief digital and information officer at CommonSpirit Health, and in previous roles was the chief of digital and tech at Kimberly Clark, Walmart, Nestle, and Kimberly and Company. Suja specializes in launching new businesses, untapped customer segments, and revenue streams, scaling profitable businesses, transforming cost structure. She's an expert in risk management, cybersecurity, digital marketing, customer experience, ecommerce, AI, and m and a. And we're really thrilled to have you back. I think you were on our opening, women in supply chain summit, and you've done a lot for women in, supply chain. So we really, are looking forward to this conversation. And I'll turn it over to you maybe to say a few words, Suja, just to frame the conversation. And then I have a bunch of questions that I've prepared as well that we can go through. Rohini, thank you. And hello, everyone. Trust you are having a a great time at this event, and thank you for joining us. I have a business tech, and people transformation leadership from the executive committees of various businesses, Fortune one to Fortune 50. I know I do a board career on public, private, and nonprofit boards. I get to see why and how supply chain has moved from a back office or even a center office function to squarely being strategic to the businesses and in the boardroom. The language of the boardroom is is strategy. We're discussing risks. We're discussing finance. And, of course, the language of the boardroom is people and stakeholder management. And, we apply these lenses to the supply chain topics. And I'm excited to share some insights on what we are looking at from this vantage view and in particular about AI. So, you touched on this, Rohini, the major events shaping our landscape, certainly geopolitical events, the tariff situations, wars, nation states. We have we're seeing extraordinary economic volatility, the upcasting, downcasting of GDP, predictions on interest rates, fluctuating demand patterns, currency fluctuations, public markets, private investor cycles. So there is quite a bit of volatility there. Consumer sentiment dipped quite deeply in February, and big retailers, Walmart, Costco, Target, all sounded some bells. And today's news, the March consumer sentiment was even worse than the February consumer sentiment. And then, of course, there is domain specific volatility. So on the one hand, you're probably seeing good manufacturing focus in defense and other manufacturing industries, perhaps chips, AI infrastructure, but then also other industries, there is domain specific volatility. And, risk management is no longer just about responding or reacting, certainly building in resilience. But I'll tell you one big consequence of all of this is that the smart scenario planning and simulations have become almost an expected norm. And and in this context, envisioning scenarios, whether it's plus 10, minus 10, plus x, plus minus x, what if scenarios what if consumer sentiment goes lower than before? What if you're going in a certain path, but the consumer sentiment picks up? What if you don't have the inventory and the consumer buoyancy has has stepped up? What is the interconnectivity of the scenarios between the various functions and groups within the company? Merchandising versus supply chain, supply chain versus store ops. What level to call? What early signals to catch? How do how how does one be smartly agile? This is a a major focus. And in one of our boards, we talked about, a a simulation scenario, and the first tranche was just doing a finance simulation. But then very quickly, the team realized that this is much more than only looking at a finance scenario, so much more holistic robust scenarios. So what's happening in the world of supply chain? It's important to acknowledge that compared to previous decades where things felt much more calmer and quieter, the black swan events have almost become the norm. It's like the decibel level of the new norm is constantly increasing, so the black swans have become white swans. And and the and supply chain means volatility. Supply chains are expected to be chaotic. So, also, these are moments of tremendous opportunities. So what are businesses doing? Every one of the companies I am associated with, there are structural focused topics. What do we do for capacity growth? How do we address service level while also looking at our sites? Our our sites of whether it's DCs or manufacturing or sourcing locations, are they adequately diversified, and are we appropriately structured for cost of these sites? The integrated planning and responding for demand volatility, I I mentioned it earlier. So this is, extremely important. So supply chains are getting rewired as we speak. The whole global supply chain is getting rewired, and each tranche of each step of the rewiring comes with challenges. But then how are businesses watching this rewiring and all the downstream suppliers to these redistributed supply chain sites? How are they watching this and applying this as a dataset and and responding to it. Inventory planning allocations, working capital, working capital turn management, all of these are in a bit of a dizzy because of the volatility, because of the extraordinary volatility. And, and automation and productivity, of course, it's it's not just about replacing manual tasks. It's about supercharging your efficiency from robotic automations to AI scheduling tools, etcetera. So that's that's what's happening in the supply chain, but then how are businesses responding to these extraordinary challenges? So the first thing is we are all talking about stacking up the s curves. So it's having a system to ideate, test, launch, scale, exploit, reconfigure, and potentially even just disband, options, but then having these s curves stack up one on top of each each other. But also building sequential s curves so that each product market revenue generation has some kind of a line and lineage to the next product market revenue generation. So how do you build that? And even if they are only transient in nature, nothing it doesn't have to be so permanent for the next fifty five years. But even if they are transient, how do you unlock that? And then breaking ecosystems. So supply chain is absolutely an ecosystem play and ecosystems within the company sometimes or ecosystem. So let's say we have a product that's selling amazingly well within a season. How do we have a change of infrastructure so that we are able to to get that? And then, of course, allocating store inventory allocation, store versus online was a thing. But instead of having an algorithm that you can now expose the entire inventory, whether it is for online or anywhere in between and have that elasticity to be able to distribute across the notes based on service level and cost. And then, of course, engaging very much with early or innovators and startups, which is very much the context of this conversation. I'm also on the board of Pandu, as you know. I see a GenTech AI in action from the front line. What, what we we create, the GenTech AI and the logistics space, but also various other companies that have implemented this and and truly driving value. And, supply chain, by definition, you're shipping atoms. You're moving products. And so I would say maybe the 70 to 80% of our supply chain is physical goods and movements, while the remaining 20 to 30% that is digitized, which is the information ecosystem, certainly is prime for adoption of AI. And, of course, robotics and human rights will have a play role to play in the physical part. And unleashing and turbocharging efficiency and productivity in that 20 to 30% is so crucial. I also think of it this way. There are macro data points that everybody talks about, and then there is micro points such as lanes, categories, the multimodality nature of the supply chains. But before you know, the intersection of all of these become millions and billions and trillions of data points. So there's no insights. Mhmm. Absolutely has necessitates the need to apply and use, AI. I I think of it this way, Rohini. My entire career, I've spent systematizing and automating. But when I really look back, I I the widening of the aperture is happening in so many different ways. So it is happening in terms of these datasets that typically you just left it as maybe a report that was read, but now it's a dataset that's part of your algorithm. Or it's a a business collaboration and an integration or a a conversation that's part of a supply chain, the phone calls, the thousands of emails and collaborations that happen between companies as products are open place to place. All of those now are are turned into data or or systematized and automated. So suddenly, now my ability to penetrate further into that automation spectrum increases, and and the aperture has now widened to almost 80% of that 30% of the supply chain. Now I can get into automation with AgenTic AI. Mhmm. And and but there's still a lot of confusion on AgenTic AI. There's a clear difference between just automation versus AI automation versus AI agents. There's a difference between a query bot versus an agentic bot. Enterprises are thinking about how do you look at system of truth or system of record versus a system of context. So all of advanced reasoning, context building, decision making, which we're building now in in in the start that I I talked about, all of that, the system of context and the richness of that context. And how do we keep all of this architecture composable and open and yet hold some kind of a a a, an ownership around the data? Right? So data as as that asset. And how do you still play in that open ecosystem but still hold a certain kind of a of a an authority over that data. And And then also the intertwining of the triumvirate. So there is the services market, which is big in any function, but also in supply chain. And then there is the space where you have the technology products, and then there's the tech services world. And all three of these worlds are coming together and doing this very interesting tango dams. And that intertwining of the Triumvirate Enterprises are keeping a very close eye, because when you look at it holistically, it is a much better profitability efficiency for the enterprise. Yeah. And, of course, I will lead that. I personally see, a AI personas, freight payment specialists, route authorizing personas. I'm personally watching this in action in my companies. So I am super thrilled about this agentic AI journey and what is to come in supply chain. Let me pause there. No. That's awesome. I think, that's a great overview, and maybe we'll jump off from some of those points. I think I was listening to this podcast on on IKEA, you know, on how the growth of IKEA, there's this three hour podcast called Acquired. They they talk about sort of how the brand promise so, you know, most retail initially started with the brand promise. It's, you know, it's this well designed, low cost, and available to the everyman. That's sort of like the brand promise. And, and then the supply chain sort of works itself out in service of this brand promise. You kinda be kinda have this view that that's how consumer retail works. But, of course, now supply chain is a lot more complicated. So have the board conversations at consumer companies kind of changed from supply chain is my brand promise to something else? Like, how has, you know, how has supply chain been elevated within board conversations in your experience? So there is it it everything begins with the customer. So what does the customer want? Why is what is our right to win with the customer in the products we serve Mhmm. And and how we play with the customer. So the conversation begins with the customer quintessentially. But then if you pull the thread of the customer all through, executive leadership teams as well as the boardrooms, the conversation is certainly around what is the best design, what are the products that we are offering that are on trend, on design, on and then what kind of innovations that's happening in the textile or or in the manufacturing process? And and then the design is then leads to engineering. What is the engineering behind? My favorite example is that a few years back, even recently, in recent past, the, the depression, the the smudgings you have in the jeans was done very manually. And that innovation moving into laser, that innovation came as a partnership between the designers and the supply chain, and and every aspect of it, the buttons, the motifs, what happens in the in the design. Every one of that has an engineering aspect, and the engineering has a manufacturing tethered into it. And also all of this has a cost aspect to it. And then also the open to buy. I've got an open to buy and a promise to commit to the inventory, and that's very deeply connected with how much demand am I for or forecasting for the next upcoming seasons. And all of this within the context of, what is the revenue, what kind of gross margin are we gonna get, much margin, and then the ultimate net margin. So all of the those conversations are intertwined. Resiliency is a major factor as a part of all of these things. Being able to chase and pulling a lever of chasing in season, that is absolutely intertwined with what is happening right this moment, which products are selling right now. And that's the right tool. Right. It's so much quintessential, into the courtroom conversation. And to your point on, these trillions of data points, like, once you once you actually instrument your supply chain so you can derive your financial results from your supply chain. You can derive your, you know, I guess you we just talked about security and the supply chain. There's so many aspects of your supply chain that you have to worry about. Is is your idea of AI at scale basically, a need at this point where you say, look, we are we're living in such a complex environment that it's not human scale problem solving. How do you kinda think about the application of AI? Where does agentic come in, in terms of the complexity you deal with, in, you know and then we'll talk about, you know, what might go wrong as well in a second. But what is the opportunity you see for AI at scale within supply chain? So the think of where we came from is even within supply chain that is systems and capabilities which are sort of tied to the organizational silos. So there is a planning set of planning systems. And then we're gonna planning all the subcomponents of planning. And then there is manufacturing tools and then sourcing. So there is some somehow also all of these are integrated interface APIs. But then as you go forward in the future world, how are all of these gonna come together as and it's not that difficult to build a knowledge graph across all of these things. I don't believe today that it exists. Mhmm. And then and will it get built category by category? Will it get built, will it get built in general? But but these are this is these are the questions that the industry will be dealing with. But then even in the short term, in the short to midterm, even if, let's say, component by component, there is good amount of automation, there is a good data collection, and you widen your aperture beyond just the core dataset that you use, for example, for for for planning purposes and, let's say, logistics planning purposes to go and and and be able to land the data of the of the lane, costs of the lane, and the fast moving datasets that are associated with that and be able to look at that and, and and manage your freight. And even, let's say, capability by capability, being able to address the data you need, internal and external. I wouldn't sweat too much that your entire dataset is, is unclean, and I was at manifest and you were, at manifest too. It was like the number one problem is like, oh my goodness, all this investment for decades and everybody all we care we are stuck is in this data problem. I wouldn't set about solving the data quality as I as the first topic, I would take it as a as a focus by focus and apply I think many of your speakers spoke about using the AI itself to get to a better metadata definition and met better data quality and then building with automation. And then look at, okay, what can we do holistically across, all of supply chain? How can I handle how can we look at holistically across the entire company? So I would still start small, Rohini, where we are. So take capability by capability, persona by persona, and make those decisions on should I am I gonna build by our partner? At this point, my point of view, a lot of the agenda is coming from the start up land. That is all these conversations around around where do we build enterprise platforms and then let people build their own agenda, or should the agenda be more tied to the domain expertise? I am this is my personal point of view. Domain expertise trumps anything else. It's not building the best enterprise architecture platforms for, for Agenctic AI. I mean, that we can get from the hyperscalers, but it's really understanding a deep domain awareness of the data, deep awareness of business problems that are being being addressed, deep understanding of how this collaborative ecosystem across companies come together and focus on building those personas, work them, scale them, take the take whatever you're taking out of it, productivity, service level, whatever you realize it, and then you work towards something that's perhaps a bigger picture. Yeah. No. Absolutely. And and maybe the I think that's that's great to having that Pareto approach where, you know, don't wait for your data to be perfect. Like, you need to start continue to run your business and improve it, over time. I think that's a great piece of advice. Maybe you have a similar maybe we'll end on that, on the risk side. So, you know, there is this sometimes you can get paralyzed by, oh my god. If I do something wrong, my whole business ends or, like, something could go really, really wrong, high consequence risks versus there are some that are not high high consequence risks. How do you how should companies think about risks as they play around with AI? And then, you know, I think maybe we'll end there. You've got thirty seconds. Yeah. We I think the tool will cut us off. This is the AI related risks is its own topic. Right? To have models and AI to help you manage hallucinations and bias and use multiple chain of reasoning and chain of decision making, architectures and approaches in your agenda to get you as close to a deterministic result as you can. I think we'll pause there. We are going to be cut off in six seconds. Rohini, it's been a pleasure. Thank you. I really thank you for the these comments. I'll see you in the on the closeout. Yeah. Thank you. Alright. Let's see.