SupplyChainTalk: How AI is modernising logistics management

Speaker A: Foreign.

Speaker B: Welcome to the Supply Chain Talk. I'm Ana Maria Velika, founder at Green Apples and I'm your host. Today I have a line up uh, of great panelists, two amazing leaders, um, and before I invite them on our virtual stage, I would like to thank you for attending our episode of on a very, very uh, important topic, how AI is modernizing logistics management. So I have to thank you for joining via the uh, business Reporter website or via LinkedIn. Our producer will keep an eye on your questions. Please, please interact with us in real time and address questions, comments, compliments. The best one will be rewarded towards the end of the episode with our fav much Supply Chain Talk mug. And um, as you know McKinsey as a powerful statistic, um, McKinsey research indicates that companies that fully deploy AI in supply chains can reduce logistics uh, costs by 30%, reduce inventory by 50% and reduce stockouts by up to 65. So AI is becoming not only a necessary, an enabler, something um, that contributes to the bottom line to the company, but it becomes like the norm. And I have two distinguished um, uh, experts and I would like to invite you on our virtual stage. Laura Storch Pelletier, all the way from New York.

Speaker A: It's great to be here today.

Speaker B: Amazing. It's good to have you here. Laura, a few words about um, uh, Laura Storch Palletier. She's um, the head of strategic supply chain for Fuku, a fried chicken QSR concept by chef David Chang and has spent her career overseeing every aspect of national and international supply chains in the food and beverage industry. Before making the pivot into supply chain, Laura went to culinary school and still enjoys cooking and baking for friends and family, especially with ingredients growing in her garden. This is beautiful and very sustainable.

Speaker A: It's my little happy place in the summer. Now is the time. So it's my little happy place.

Speaker B: That's beautiful. Um, our second um, panelist, Cindy Brandt. Cindy, if you can uh, turn on your camera. Welcome to our uh, Supply Chain Talk.

Speaker C: Cindy, super excited to be here.

Speaker B: Yeah, it's good to have you here. Uh, a few words about Cindy Brandt, uh, vice president, uh, of industry solutions at uh, Descartes. Cindy has spent more than 25 years last mile tech watching it evolve from maybe computers can help to full scale digital transformation. Powering everything from planning and dispatch to execution, customer communication and real time driver, vehicle and safety insights. And Cindy is just getting started with AI, automation and predictive intelligence set to redefine what's possible next. As an interesting fact, a Customer once dared her team to ride a real bull. He was a pro on the senior circuit. No pressure. And everyone suddenly found the urgent reason to opt out. Except Cindy. And um, 2.5 seconds later she was alive. Can't contract one. This is impressive.

Speaker C: Anything. Anything for customer. Anything for a customer.

Speaker B: For a customer, exactly. I'm with you. Um, I would like to brought to um, our audience attention a um, breaking news um, item, um, linked to Amazon. Uh, why it's breaking news because they are daring greatly, I would say they are daring greatly to bring robots in, um, making our experience as shoppers like exceptional. And uh, you have on your uh, screens now the link uh to the news item, um, which is published by Reuters. Amazon unveils new AI warehouse robot in US$12b Europe push a few highlights. Uh, Amazon, the Seattle based e commerce giant showcased the next generation Proteus robot at its Delivering the Future event at its Dartford fulfillment center east of London as it works to speed up deliveries. And at this fulfillment center in Britain they unveiled this robot, an upgraded AI powered mobile for its warehouses that can respond to conversational prompts as part of 10 billion uh, investment in its European fulfillment. And this new version uh, which is due in Europe in the first half of next year can operate across warehouse floors and marks a shift in how employees interact with robots. Uh, the workers can order the robot, what needs to be done. The robot figures out the priority, the route, the timing. So this is quite exceptional. And the ambition, Amazon ambition is same day delivery for fresh groceries is now available in more than 2,300 in the US cities and parts of Tokyo with further expansion. So this is massive. I would like to invite your thoughts Laura on this breaking news and Amazon daring greatly.

Speaker A: Yeah, well I think to your point it's another example of a key industry player making a huge move into AI and really hinging its growth expectations on what AI is going to be able to do. Coming from the food and beverage world, the grocery aspect was especially uh, interesting to me. Um, you can imagine how route optimization using seasonality, traffic patterns, real time traffic patterns, all of those types of things is going to be really transformative in the ability to do same day delivery and really expand. You know it's, it's definitely something that's going to be continuously expected from customers I think for more and more commodities and we're going to see more and more investment um, mirroring this sort of thing.

Speaker B: Thank you Laura. Um, Cindy, what are your thoughts?

Speaker C: Well, I'm always going to look at this a little Bit from a last mile perspective. Right. So I look at this article and to me it's not really about warehouses, it's about compressing that entire order to delivery timeline. Uh, and they really are trying to help reshape once again last mile economics and expectations. I mean we reshaped them probably three or four times in the last decade uh where Covid kind of pushed everything into uh, a more accelerated time frame. But now as we have more information, we have more expectations, we're constantly wanting more options for delivery and faster uh, you know it's, it's really interesting and as a carry on to that, you know this is upstream automation to enhance um, last mile cost and speed. Right. I mean that's what at the end of the day that's what we're really trying to do. The robot is obviously going to improve pick and pack speeds, uh, you know, and like Laura alluded to that route optimization inside the, the uh, facilities. But that's going to help them increase uh, order cutoff times. They can do more, right? They can do it more repetitively. Second they can do kind of these late wave releases so they can hold the uh, ordering time open longer and do that later cutoff time and it's really going to allow them to do a much denser kind of last mile uh, routing window.

Speaker B: Right.

Speaker C: So shorter windows, more stops. Uh, you know I love this because it's really attacking kind of a lot of the last mile constraints way up high at the source. So they're just trying to prove that faster fulfillment is going to drive these later cutoff times higher delivery density. Uh, and that's where you start to see the real cost savings, especially in delivery density.

Speaker B: Yes, it's exciting and um, we all as shoppers we will benefit from that. I'm really excited about Also they announce Alexa plus next year. So let's see what uh, the future holds. Um, we go now into our first topic, um, about transforming logistics and fleet operations through AI. AI is changing our lives, how businesses operate and is forcing businesses to transform from within to drive smarter, planning optimizations and performance. And if I remember my times at Nestle, uh, when I was leading supply chain for uh, their nutrition business, we didn't call it AI, uh, we call it planning excellence. Because the biggest challenge wasn't moving uh, infant um, products, formula and uh, food around. It was predicting demand accurately enough so that the right infant formula was available in the right hospital, pharmacy or retailer at the right moment. And every single error, if you can imagine in this critical product um, of Only few percentage points could create shortages, waste or unnecessary transport cost. Today AI can process hundreds of variables simultaneously. Events, how we call it in demand planning like weather, promotions, epidemiology data, social media trends or black swan, even data, uh, in order to generate demand signals. Humans can simply cannot compete with this. And I would like to invite uh, Laura, from your experience, if you have some examples in, in this way, uh, transforming logistics and fleet operations through AI.

Speaker A: Well I think what you said just now really resonated with me where it's really a planning excellence tool, um, and that's really my main capacity in using it. You know, I've done 200, $300 million supply chains on spreadsheets and past years sales data and you know, hoping for the best. I think of just you know, 10 years, 15 years later and now I'm able to do real time cogs analysis where I'm using, I'm taking the items that I'm moving as a customer and seeing where are these cogs coming in based off of, you know, commodity markets, where can I be saving money? I'm able to now connect to our distributor warehouse management system. You know, as a restaurant group we are using a broadline distributor to do all of our last mile deliveries. So I don't have as much uh, visibility as I would with certain other arrangements. This bridges that gap and helps with the planning of that. Now I can incorporate expiration in a more perfect way. Expiration dates, production dates, things of that nature, um, and then also connect it with our erp so, so I can see, you know, there's a lot of seasonality, we're in a lot of concessions, you know, just naturally restaurant peak times, um, and I'm able to now incorporate all of that at the same time in a way that a human brain um, is not going to be able to do all at once. So it's really transformative for us in that way. And I think we're only really beginning to touch the surface of what it will be able to do for us. But just to see how comprehensive it is, um, is a huge game changer for us.

Speaker B: Exactly Laura. And what you mentioned, you mentioned about working collaboratively and jointly with your uh, customers, distributors, uh, and this is key, it's about reshaping, retransforming the internal processes in order to get the right outcome. Because AI can give us so many outputs. But if we are not transforming the way we integrate, we jointly do forecast together with our customers, retailers, uh, AI is just a tool. What's your experiences in the end from your experience.

Speaker C: You know, I love that you guys are actually talking about, you know, planning excellence or that planning process. Right. Because when you think about last mile, there's so much noise in the market right now about, you know, AI based routing. And the reality is, is that it's really using AI to look at your data and make your data cleaner and more accurate so you can create more accurate plans that are much better at being predictive and right as opposed to best guesses. So when you think about last mile systems today, a lot of things that we put in are best or educated guesses specifically around things like how long it takes to service a customer. Uh, you really have to harness a lot of that information on the execution side to make that planning piece much, much more accurate. So thank you so much for both of you for talking about the concepts of planning because I think that's something that's really missed in this conversation around AI. But when we step back a minute and then look at the bigger picture, Last Mile transportation and logistics on the whole is a lot of times about reactive logistics management. Um, we're now using AI to make that transition to both predictive and adaptive operations, if you will. So instead of responding to something after it happens, you can use AI and AI agents, actually anticipate that disruption and adjust the routes, rebalance capacity, or even make communications to external constituents, whether they're the customers or your own internal salespeople, to uh, adjust for that, that delay, if you will. So you know, I love the fact that people are now taking the time to think about the workflows and the outcomes and how AI can not eliminate people in the supply chain, but automate highly repeatable tasks and free up the human beings to make what I call the gray decisions. You know, AI is going to make those black and white decisions, but humans will, are allowed to make that gray.

Speaker A: Very well said.

Speaker B: Exactly. You will uh, need those brains, uh, experts, uh, bold enough to understand all the scenarios. Exactly what you are saying Cindy, because the AI enable us, uh, so many scenario plays with so many risks scenarios in order to focus on the right solution, uh, in the most effective way. Um, and if you don't have those people and that leadership ethical courageous to speak up and to go for, to press the button. I think um, AI is just AI. And you mentioned about logistics and um, routes optimization. We know, uh, in my time at, for example at Heineken, uh, AI was driving route optimizations, uh, which um, uh, delivered like 50%. So they cut delivery routes by 50% can you imagine the economy of scale, uh, and the benefits, the financial benefits, not only the time spent on the road for drivers, but also optimized cost.

Speaker C: There's so many operational ripple effects from using AI and it's everything from driver retention and all the way through cost savings through increased density that allows you to grow your business without having to add more assets. I mean, it's just the possibilities here are really, uh, tremendous for every organization that deploys AI.

Speaker A: Absolutely. And for an emerging brand like Fuku. You know, our ability to scale is our ability to deliver, and our ability to deliver is can we get your product where it needs to be so that you could have an operational restaurant at the cogs you need it to be profitable. Um, and AI is making huge hedgeway in the, in the way we're able to guarantee that to be able to grow.

Speaker B: Yes, Laura, we have a comment, um, from the audience. Uh, thank you, Charlotte Spelling. Uh, automated fleet management is the big hush hush. If no one is really sharing strategies, models, outcomes, competitive advantages, then how can I measure my organization's maturity? Are we the early adopters or behind the curve?

Speaker C: Cindy, that is a great question. Uh, and I think the answer is really working with great vendors, right, that are actually looking at that information. Uh, you know, we collect a lot of information, but obviously we keep it in separate buckets. However, we can have conversations about generally what we're seeing. Um, the other thing too is, you know, working with uh, the different trade associations, you know, we've been working with them to understand how do we create some good benchmarking studies that can be shared across things. But I think that this is, I uh, truly believe that the transportation ecosystem in general is always hush hush because people think that, um, you know, there's secrets, uh, if you will, you know, there's some secret sauce in there that they don't necessarily want to share. But uh, the other thing is having conversations with what I call non competitive similar businesses. Right? So for example, if you're a food distributor, have a conversation with a beer distributor, you know, think find a parallel industry, but not your exact competitor. Um, more often than not people will truly have wonderful conversations with you. And then the other thing too is make sure you go to your vendors, user conferences, there's so many people. And trade and trade conferences, I should say that too. I don't want to, um, pigeonhole it there because you know, I always say the point there is to network, not necessarily, um, go to all the sessions, but who can you talk to and network with? So you can do exactly this, understand. How do you benchmark against other people? How do you create that network? Uh, it's really important.

Speaker B: It's very important. And this brings me back to time, to the Brexit time and Brexit was um, a tough teacher. Um, and I remember in Britain it was like a forum on a monthly basis like the Food Nation, so Feeding the Nation, um, Crisis committee and a selected panel of corporations in the food and beverage uh, were there. It was quite scary to sit at the table with politicians and uh, people um, from the economic, you know, from the political and economical scene in Britain. Just because I was in charge with a uh, critical. Bringing the critical product, importing the critical product for babies in the UK all the way from the US because uh, there were only two factories, one based in, in the US So we had to, to report uh, and we had, it was such an open forum, Cindy. Uh, we were sitting next to our competitors, um, Danone, who was also importing and manufacturing the same uh, formula. But sometimes uh, as you say, um, we are stronger together without anti competition, without going to the anti competition, you know and breaking any anti competition law. We can find ways to push this AI in an ethical way. And because you mentioned uh, the two of you about um, how important is this integrated way of looking at processes at data, I would like now to go into the second topic, leveraging data, ah, automation and visibility. Why it's important. It's important because we are here to serve our customers and our consumers. And there is a uh, statistic brought by McKinsey, a research showing that visibility initiatives could reduce stockouts by 50% and emergency freight costs by 30%. From your experience, do you have an example uh, Laura, about all these three buckets?

Speaker A: Yeah. I mean just in the beginning of being able to incorporate AI into our planning, we're seeing 10% in some key items, stock out reduction, um, and now being able to incorporate items that we had maybe less visibility in as non proprietary items. The AI is able to gather all of this information, give us more, much more comprehensive information. So we've seen a lot of decreases in that. But then also we are able to use industry data to benchmark this and decide when should we be buying more, when should we be holding off. Um, so it's long term planning and it really touches into everything. Um, especially with now managing expiration dates for short shelf life items. Um, it's been transformative for us in that way for sure.

Speaker B: Yes, um, exactly. And you need also to have all these scenarios for various uh, um, various um, situations, um, in logistic management, do I use um, train, Do I use uh, um, I don't know, air freight? Um, is it more important to use air freight, although it's uh, five times more uh, cost? Um, yeah, more expensive but probably it gives me availability of the product. How are you, uh, from your experience Cindy, uh, at Descartes, um, how do you balance the data, uh, automation and real uh, time visibility?

Speaker C: Uh, I think for us, you know, the major theme uh, in last mile transformation right now truly is all about visibility, right? And visibility at several different, different levels. Where's my order? How are my drivers doing? How are my vehicles doing? Uh, you know, how is my business doing? Right? And if you think about the customer expectation, they're starting to really expect information about accurate time windows, super near real time live updates, right? And of course proactive communication. If there's an accident on the freeway and you're going to be 15 minutes late, I need and want to know that because there's a lot of again ripple effects through their daily operations. But if you balance that, you know, at the very same time logistics needers really uh, need a clear view of both fleet performance, uh, driver activity and driver safety. I would even say, you know, where are the bottlenecks within the entire process? And of course where are these exceptions that kind of get thrown out in the network as well. So if you can connect AI to high quality real time data, that's where that difference is so significant. And if you think about the data we're trying to connect today, it's vehicle telematics, whether it's driver and vehicle behavior or cameras, uh, warehouse systems, customer order data, uh, let's see, inventory platforms, uh, uh, information coming in from mobile driver workflows in the field. And of course you know, how well a route has performed uh, historically and even historical order volume.

Speaker A: Right.

Speaker C: And then I, I would dare to say there's another data layer in there right now that becomes important when we're talking about reducing cost. And that's going to be safety and how safety parlays into things because uh, you can drive down transportation costs, but you can also drive down insurance costs as well. But when all these systems work, you know, together, it takes all these fragmented processes and unites them, right? So now I can start to actually report both internally and externally on what's my on time performance, you know, on time, uh, in, in full, uh, you know, what's uh, the actual cost per delivery. So I know is that customer really generating profit for me um, you know, or do I need to do some consolidation, uh, you know, and look at delivering one day a week instead of two Driver, uh, productivity today. We always have a challenge with drivers which is either they're too overloaded or they don't have enough work. And that's because our data isn't accurate. Right. So using AI to understand those service times, uh, and then it ripples down again to like fuel, um, um, insurance costs, uh, service reliability, a lot of different pieces there. Um, I think you're going to see the strongest use cases uh, when you combine like I was talking about before all this automation with the human judgment. Because AI can make those surface decisions, um, and recommend some actions. But your logistics team really has to dig in sometimes.

Speaker B: You need credible expertise and this is the role of us, ah, at the table and this is what we bring back. AI will enable stronger data, stronger scenario play, uh, with stronger financial business case. But it's nothing without a credible expertise at the table to influence and to challenge the C suite to take uh, the right uh, decisions. We have a question from the audience and I would like, uh, Laura if you can try to reply. Uh, AI needs high quality, real time visibility, open data, uh, streams across carriers, shippers and third party uh, logistics providers. How do you convince these highly competitive carriers to share their data, uh, with shared visibility platforms?

Speaker A: I mean I think as it becomes more and more of an expectation I'm seeing a little bit less and less resistance. Um, it's also something that I've been keeping in mind as I select vendors of wanting that transparency. It's been something that I've been you know, front of, uh, conversations to make sure that we are getting that because at this point it is essential, especially for our setup and not controlling the last mile in that way. We need to be able to have the integration to be able to speak to everything together. Um, so I mean in general my philosophy, uh, when it comes to vendor management is to try to you know, build up relationships with a couple of key vendors and leverage that relationship for situations like this. Um, so that would be something that I would, I would definitely recommend. And also just seeing, you know, in your particular sector of commodities and transportation and logistics, what platforms might be there that might be able to kind of consolidate. Um, you know, if they don't want to speak to you directly, I'm seeing at least in, in our world, um, different platforms that they can integrate directly with the warehouse management system. Even if they don't want me as the customer getting it. You know, I Kind of get it filtered in that way. So sometimes it just takes a little creativity. Um, but I think it's something that it's good to ask for. And the more um, these vendors hear it, the more likely they're going to be to share this information in the future.

Speaker B: Exactly, Laura. And I think it's also a creativity and uh, an education thing. Uh, because they expect all these competitors, uh, they expect us to educate them on some platforms. Some, some of them from my experience, they don't have the know how, they don't have even the internal resource to allocate. Okay. So then they say no, no, no and they push back. Ah, but as you say, A.I. is here, um, for all, uh, A.I. is here to stay. Uh, so it's becoming a normality. Um, we have also a question, uh, so keep the questions coming. I love them. Uh, questions from Tom Hold. We are constantly balancing service levels against transport costs. Where have you seen AI make the biggest difference without creating extra complexity for the operations team?

Speaker C: Cindy, that is a great question. I mean, and that is also the age old balance, right when you come to transportation. Uh, so what we've seen is that one, people make educated guesses in their planning systems. So they'll say, I'm delivering to a bar, I'm going to allocate 22 minutes. Well, the reality might be that it's a bar in downtown historic Baltimore where I don't roll through the front doors, but I have to open up hurricane doors and take it down a ladder to the basement. Right. Totally different service time. So that's my example. But the reality is that we have to understand the preciseness of those service times because one, we may be creating routes that are 100% unrunnable. Which, the ripple effect which is my theme today is going to be that you're not going to retain drivers, you're going to lose some tribal knowledge along the way. But if I, or conversely the routes are too short, but your drivers don't come back until eight hours because they get paid for eight hours. So they might be able to do other additional stops as well. So that's where truly understanding what is the service time per individual location. Right. And being able to predict that service time and potentially even as you add new customers, being able to predict what their service time will be until you get that accurate information collected over time from servicing them. So I think that the reality is that we have to take a hard look at are our routes runnable and that's what AI allows us to do a lot of times is look at all that data, model different scenarios to actually understand should my service time for this stop be 22 minutes or 14? Because on average it might be that much shorter time. And if you're lucky, which we find many of our customers are, uh, we're seeing, you know, significant improvements of essentially finding excess capacity within the routes. Uh, so they're resetting expectations. But uh, the other thing too is making sure that the routes are planned in a safe manner. We don't want to rush anyone or compress the route. That becomes a stressful factor, increases the chance of risk, increases the chance of accidents. We want to make sure we're pulling all the information together to create the most optimal routes that benefit the business, allow the business to grow, but also allow the drivers to run the routes and be safe.

Speaker B: And be safe. Yeah. Um, I remember from my time at Nestle, um, safe and zero accidents was a must. And for any publicly listed company, uh, this, it's a public commitment, um, safe, uh, and safety comes first, especially in the manufacturing environment. So I, I couldn't agree more. Um, and for me personally, if I speak from my experience over the last, uh, 20 years I've helped move everything from infant formula, infant, um, nutrition at Nestle to beer at Heineken, and tobacco products at bat. And for me, one lesson remained true. Supply chains, they don't fail because products they don't move, they fail because decisions don't move fast enough. Because companies are fast at implementing these platforms. Um, but they are very, um, they don't recon. Recognize the importance of capability built. You bring AI units, capability built. And the question is no longer whether AI will transform logistics. The question is for me, in which organizations um, will transform themselves quickly enough to capture the value. And I would like now to go into our last topic, which is about turning digital ambition into measurable impact, which is translated in leadership, governance, culture and technology partnerships. Um, what are your thoughts, uh, Laura?

Speaker A: Um, well, you know, if it was up to me, I look at it and I think you guys spoke to it well before as well. I see AI as an incredible tool. Um, but I don't think it is a replacement for supply chain and quite, you know, the way that certain people, sure, certain tasks, certain things will be automated, but there's so much that comes from person to person conversations, um, not even just in negotiations, but just in those relationships and how it builds and understanding those nuances. So you know, as we continue to push, you know, and also just hoping for stability, uh, all of our companies benefit when there's a consumer base that has money to spend. So we want people gainfully employed and you know, everything chugging along in that sense. So I hope we continue to use it as that tool, um, to further aid a robust working network.

Speaker B: Yeah, this is beautiful. Cindy, we have a question before, uh, I invite your thoughts on this topic. Um, from Jessica. If you were advising a mid sized uh, logistics operation today, what's the one AI use case you'd prioritize first and why?

Speaker C: So my first answer would be a question back, which is what's the biggest outcome you're looking to change? Uh, you know the, the, the, the challenge was answering that question directly. And I will, I will though uh, is that every business is a little bit different. Right? Um, we have seen the most opportunity in looking at our data sets in service time. So I would say looking at accurate service times and here's why, when I put bad information into the system, bad information comes out. So you know, use AI to make your data smarter, more accurate, uh, and more usable. So that, that's what I would say. And you could use AI to normalize customer addresses, um, analyze service times, analyze where they're actually located at geocode, which is, you know, the lat long, which is important for last mile optimization software. Uh, we said open, close times, time, windows, service time. These are all important facts of data that is often collected in four different spaces and different people have different opinions. Salespeople uh, always want their customers to have the time window first thing in the morning. But let's face it, everybody can't have the first thing in the morning delivery. Uh, it just doesn't work. So it's a great question. Um, that's what I would prioritize if it was my business. But I do really think you have to sit down in your business and understand where is the biggest cost pain because that's what you're trying to uh, eventually solve with AI from my standpoint is improving efficiency, productivity which reduces your overall cost.

Speaker B: This is beautiful. And I would add for Jessica, look at the type of category, what kind of um, industry you are operating in. Because if you take two products, indulgent and critical, uh, infant formula and chocolate, I worked for both. Yeah, uh, which, which one you would prioritize in terms of uh, AI? Definitely the critical product over the indulgent one. Also probably the indulgent one has a higher uh, margin. So these, all these considerants, all these considerations needs to be uh, looked at. Uh, so you have already a master class. I hope you took Notes Jessica. Um, and uh, Cindy, on our topic about turning digital ambition into measurable impact, leadership, governance, culture. What are your thoughts, Cindy?

Speaker C: So I think that uh, it's interesting because I feel like every M Trade show I've been to recently or every talk, uh, they ask everybody in the audience how many people have started an AI initiative. And everybody raises their hand and then they say, well, how many people have had a successful initiative? And like two people raise their hand. Right. So I think that there is a real problem here. We all think about AI, we want to use AI, but, but moving it to be impactful, um, is difficult. So I think one, you've got to start with really understanding the business problem you're really trying to solve. What's the workflow today? You really have to sit down and talk to everybody that touches that process so that you can figure out what happens and where the highly manual, repeatable tasks are. Because I can take an AI agent and get that manual task and automate it super, super fast. That's a really easy thing to do. Two, you have to have data governance. Uh, the theme that all of us have talked about is bad data equals bad answers. And even from some of the questions we had today, how do you get that clean data? Well, you can use AI to help clean your data. Don't forget that. Um, the third thing is really about cross functional ownership. I can't push an AI initiative in that's going to touch multiple people and not have it, not have the buy in from everybody. Right. So if I use AI to deal with fleet exceptions, I have to make sure that the customer service folks are going to be able to use that, uh, as well. Um, and then the last and probably the most important thing is, uh, buy in from your frontline teams. Right. If you're going to deploy technology, have a conversation with the people who will end up using it. One, you're going to learn something you didn't know. But two, if you get the buy in on the front end, as soon as you implement and deploy, you're going to already have earned the right to deploy and earn the right to change their process. And I think Laura, you've probably dealt with that a, uh, hundred times too. I mean it's just so, so important to have the conversation.

Speaker A: Yeah. And just pointing out all of the efficiencies and all of the benefits that come from it to make sure all the cross functional teams are understanding how this will impact the business for the best overall.

Speaker B: Yes. Beautiful conversation. Uh, which could carry on and on it's just that we are, it's unbelievable. Uh, we are approach reaching uh, very quickly, rapidly the end of our talk show. Uh, thank you to our audience for making it so interactive, um, and for commenting and asking questions, uh, in a very uh, subjective way. I will choose uh, one, uh, winner and I would like to offer our supply chain, uh, talk mark to Tom hold, who addressed the beautiful question about uh, constantly balancing service levels against transport costs and where we see AI making the biggest difference. Thank you Tom. And uh, watch the post. Um, and because my colleagues at Business Reporter will courier the, the mug to you and before we um, wrap up, I would like to invite you Laura, for your one closing thought and action item for our audience.

Speaker A: I think building on what I said last, I see this as sort of a two part development. We're simultaneously building up those relationships and being very transparent with our vendors, how we're incorporating AI while we're going through those processes on our end and having the full capabilities of AI cleaning up our data. To Cindy's point. And really as we get into the demand, forecasting is where we see um, the biggest transformation happening.

Speaker B: Beautiful, Beautiful Laura. Thank you Cindy. Your action item and closing thought.

Speaker C: So my action item is to stop thinking about using AI, uh, and start to do the foundational work so you can use AI. So really start to look at what do you want to change, what outcomes do you want to drive, and when you identify those outcomes, then start to really look at the workflows and uh, you know, don't be afraid to ask questions about your vendors and what their AI roadmap looks like, uh, and how you can apply that to your business and ask them the hard questions on how it will impact their business. Right. So you know, proactively create your foundation. Number one, what do I want to change? Uh, do the work to understand the workflows and two, have the conversation with vendors and ask them what they're doing and specifically how it's going to impact the business.

Speaker B: Beautiful. From my side, uh, it's a question like a mentoring or coaching question for the leaders who take this initiative to implement um, um, AI. Um, is AI fundamentally a technology transformation or is it actually a leadership transformation disguised as a technology project? Be prepared to be challenged to uh, challenge the status. Vo be prepared to equip and invest in capability, build. Be prepared for your it to be, um, scrutinizing this AI in the first place. Uh, thank you. Thank you Cindy. Thank you Laura. It was a pleasure to have you here. I will be delighted to have you. Again, I think the audience really enjoyed and we have compliments and many thanks, uh, for your answers, for your expertise. It's a master class, how I like to call it. And, um, I would like also to invite our audience to dial in. Um, our next Supply Chain Talk is on the 1st of July, uh, at 4:00pm UK time. A beautiful topic hosted by Alester Chartan, my colleague. Ah. From forecasting to Frank foresight, Rethinking demand planning in response to rapid market change. Thank you so much. And make sure you stay healthy and safe. All the best.