Sustainability doesn't cost money. If you can think about things in terms of a more holistic approach, sustainability could and should actually save you money and make you more effective, and efficient in the long run. I mean, efficiency and effectiveness have long been the foundations of a successful supply chain and that's exactly what drives sustainability. I just think that sustainability sometimes has a bad rap of being a PR tactic, or just for regulatory requirements. Good morning, good afternoon, or good evening, wherever you are in the world. Welcome to episode 65 of the Sustainable Supply Chain Podcast, the number one show focusing exclusively on the intersection of sustainability and supply chains. I'm your host, Tom Raftery, and I'm thrilled to have you here. A huge thank you to this podcast, amazing supporters. You make this podcast possible, and if you're not supporter but you'd like to join this community, support starts at just three euros or dollars a month, which is less than the price of a cup of coffee. You can find the support link in the show notes of this episode or at tinyurl.com/ssc pod. Now you know how supply chains today are being pulled in every direction. Cost pressures, sustainability targets, AI, disruption. And everyone's trying to figure out how to actually make progress without just adding complexity. Well, my guest today, Jen Chew helps some of the world's biggest companies do exactly that. She's an expert at turning messy tangled supply chain challenges into real opportunities for efficiency, resilience, and sustainability. Today, she's here to share how AI is moving from hype to real use cases in supply chains and why sustainability doesn't have to cost more, and what companies need to get right if they want to stay ahead. But before we get to that in the next few weeks, I will be chatting to JF Gagne, who's the Chief Strategy and Product Officer for Pendulum, Don Weatherbee, who's the CEO of Regen X, Danny He, CEO of Soapbox, and Jim McCullen, CTO of Century. But back to today's episode, and as I mentioned, my special guest today is Jen. Jen, welcome to the podcast. Would you like to introduce yourself? Thank you. It's great to be here. My name is Jen Chew. I'm the Vice President for Solutions and Consulting at Bristlecone. We're a supply chain services company and we help global 2000 companies deal with planning, procurement and logistics challenges. And really excited to be here with you today. Fantastic. And what was your journey to where you are today in Bristlecone Jen? Oh it's been a long one. I've been doing supply chain and advisory work in one capacity or another now for 30 years. I started doing SAP work with Pricewaterhouse before it was even PWC back in the day. So now I've completely dated myself in all sorts of verifiable ways. But yeah, it did start with With SAP as a foundation, like so many of us in supply chain have, it really seems to be a, a common backbone for so many large companies. I also, asked exactly. I spent time at Forrester Research leading their enterprise applications research streams. So writing about SAP and Oracle and also writing about manufacturing and supply chain as one of the focus areas because it was such a big part of the ERP focus area. Most recently I spent eight years at TCS in their consulting and services integration practice and leading a big transformation effort there helping TCS to become more consultative in its interaction with its large clients. And so that intersection of supply chain and consulting and transformation really was the reason why I ended up at Bristlecone, 'cause that's exactly the type of services that they provide to, their clients. Interesting. And all along that journey, what do you think has surprised you the most? That's a good question. I think it's interesting how often some of the same challenges keep coming. up over and over again, whether or not the specific technology changes or the the news story of the day that gets everybody wound up. It really all comes back to you've got to address the people, the process, the data and the tech issues in collaboration with each other, dealing with one at the expense of the other just never works out. So I guess it's surprising to me that 30 years into it, and I'm still dealing with the exact same set of challenges. So probably means I must not be a very good consultant after all, if we're still solving things 30 years later. But I think it's just a universal set of challenges that we've got to continuously, try to overcome. And Bristlecone, it's an interesting name because the Bristlecone pine is one of the oldest living organisms on the planet with some of the Bristlecone trees being over 4,000 years old. Is that related to the company name? Is like that a subtle nod to resilience? Is it like an apt metaphor for building supply chains that stand the test of time? Or is that just a coincidence? No, it's not a coincidence at all. It was a, a very purposeful naming of the organisation. It's as you pointed out very old and very resilient. It survives fires. It's an amazing organism. And I think you're exactly right that it speaks to the resiliency of our organisation and the resiliency that we try to support our clients achieve. Let's dig into the the subject matter for today's podcast. We're gonna be talking a little bit about you know, obviously, supply chain, sustainability and AI, I think is probably the topic du jour really. In your work? Have you seen a shift in AI discussions from education, bringing people up to speed and what it is, towards more implementation? And if so, what does that look like on the ground for companies who are rolling out AI solutions in their supply chains? Yeah, that's a really good phrasing of the question and exactly the type of transition that we see AI going through. Conversations a year or two ago were all about education, especially in the executive suite. Every CIO, every Chief Supply Chain Officer needed to be able to understand what AI was and be able to explain it to their peers in the C-suite. So there were lots of small experimental projects. And often they were assigned a little SWAT team in the IT organisation to go and figure out what AI was. What we now are seeing increasingly, is that AI has become more a challenge around what do we implement? How do we implement it? What do we start with? So I think folks have gotten over the hump of the basic understanding of what generative AI is. And we should talk about the difference between AI and generative AI, 'cause I think that is one part of education that hasn't really broken through yet. Sure. Folks are now shifting into the how do I do something with AI that is meaningful and has value. If I pushed a little further on that, I would say a year ago, I never heard the phrase use case associated with AI. It was more experiments, learning, right? And so people were just kind of throwing various AI tools against the wall and seeing what stuck. And now they're being much more orchestrated and structured in their approach to thinking about AI. So one of the things that Bristlecone has developed is a series of 78 supply chain specific use cases, and we're systematically running through those 78 use cases and deploying them both internally as as examples or specifically with clients. They're real and meaningful and span everything from, analysis through implementation and testing and even tackling areas as large and complex as sales and operations planning. And that's a conversation that we weren't having a year ago for sure. And seeing as you brought it up, talk to me about how you define the difference between AI and generative AI, because AI is a very broad term encompassing all kinds of things. So yeah, walk me through your, your interpretation of that. Sure. I think in just casual conversation, when somebody says AI, what they really are thinking about is generative AI, right? The stuff that hit the news what, two years ago now with ChatGPT. So AI quickly became shorthand for computer that does something that looks like thinking. Whereas AI at its most basic really is just, it It has some sort of automated learning component to it. And so AI has been around for 50 or 60 years in terms of a machine learning capability. It hasn't been able to write stories or create poetry or, really have a human-like conversation until we got to the generative part, but it is important that when we're we're talking about deploying AI in a company's supply chain, which capabilities are we actually talking about? Are we talking something that's programmatical machine learning? Or are we actually talking about generating net new content? They're both applicable. They're both important, but they do drive different use cases, different capabilities, different tactics. Yeah. Yeah. I mean, obviously things like autocorrect is very simple AI. Or, the ability of digital cameras these days to isolate faces and focus on them specifically, that's AI as well. So yeah, there are all kinds of computer vision, natural language speech processing. The spectrum of different types of AI as I said at the start is, it's enormous. So yeah, I can see and they'd all have different potential use cases in supply chain. Generative AI I can see some great uses for that in things like call centers, for example. That's, that's one great use case for it. I'm sure there are hundreds more, but that's that's one I was listening to a podcast about that this morning. So that's just why that came to mind straight away. I was using copilot the other day to help me co-author a, a white paper on tariffs, right? And so I didn't just turn the whole thing over to copilot and expect a white paper to get generated, but I did iterate over and over again in terms of laying out the agenda, starting to populate some basic content in different paragraphs. That's something that required some interesting skill sets on my part on learning how to iterate with,, with copilot, being conscious of the type of insight that copilot either was able to or limited in what it could generate on, my behalf. So those are, you know, all sorts of interesting use cases with generative AI specific to the supply chain now. Yeah, I often use the different generative AIs to fact check each other. So, for example, if I'm in ChatGPT and I'm asking it to help me craft a bit of text. If it comes up with some fact inverted commas, I generally take the fact over to Copilot and said, I read this somewhere. Is it true? And if so, gimme the sources, those kind of things can be useful as well. Yeah, interesting. It it doesn't even have to be factually incorrect. It could just be incomplete. So, for example, this tariff paper that I was working on, right, and I asked it to pull up some of, of the news articles about the sudden resolution between the United States and Mexico. And it gave a, you know, a very economic textbook kind of answer and it really didn't bring up the fact that the real resolution was Mexico agreed to put some police on the border. What copilot told me wasn't wrong, but it was incomplete. So you're exactly right. Companies do have to understand the data sources and have processes in place to validate and check and be really confident in a way that we didn't necessarily, we always have had to check and validate and be confident. But when it was, you know, your supply chain planner, you had a different level of understanding about their background and, you know, the processes they used. Generative AI is just still a little bit of a black box for most of us. And hard to understand and trace back how it got to some of its answers. So, big risk there for organisations to turn too much over to it at this point. True, true. And with all the buzz around it, how do you separate genuine opportunities or use cases for transformation from the kind of over-hyped features that might not deliver real value? Well, the good news is that all of these more sophisticated use cases require a lot of the same basic preparations. And so, while some of the noise, right, is settling down and we're trying, you know, we can start to have some early experimentation to show where there really is value. The prep steps all need to happen and companies can go ahead and be investing in in those now while they wait to see which early bets are gonna pan out. So what do I mean by that? There's a series of culture and training and skills that need to be developed for AI to be leveraged at scale, regardless of what use case we're talking about. There's a series of of data readiness activities that need to be undertaken regardless of what use cases, right? We're eventually going to be talking about, right? If you've got lots of different versions of your vendor master in your system, it would be really hard to be confident in letting machine run you know, a strategic sourcing or a contracting program within your procurement organisation, right? So there's people readiness. There's data readiness that are all good investments regardless of which specific use case. Then to actually answer your question, right? Which use case we're talking about, right? It's a matter of the same sort of prioritisation that we've done back when we were trying to figure out where we were going to do any customisation around our SAP deployments. What's the effort and what's the benefit? And so things that were lower effort, higher benefit you prioritised. No magic here, right? We have to do the same sort of thinking around sequencing the AI use cases. I would say, though, one of the differences is because of the amount of underlying prep around people and around data that need to take place, that anytime that you can start to bump to use a, a not attractive term. But any, anytime you can start to group those use cases in a particular function or a particular department, you can essentially pilot in a more effective way than having a scattershot approach. And having, you know, half a dozen use cases that span planning or procurement, or logistics, or manufacturing and not necessarily getting enough volume of insight to really be able to apply at scale in your organisation. It is both the the old fashioned, you know, what's my risk and reward and effort? And is there a way that I can group them so that I can have some real learnings in my organisation before I start to deploy more widely? Okay. And you mentioned sales and operations planning a while ago as a potential use case. So talk to me a little bit about AI's role in SNOP for, I don't know, better collaboration and scenario planning. Do you have any examples of where AI driven SNOP changed decision making for the better and how that impacted sustainability outcomes maybe seeing as this is the sustainable supply chain podcast. Yeah, so I actually see SNOP as one of the most requested use cases for companies that are on that early adoption wave because SNOP is time consuming, uses a lot of data, right? And 'cause we're talking about global data across multiple processes across your supplier and distribution network. And there's a lot of probably thousands of decisions that go into every SNOP cycle. So we're not at the point, nor do I think we ever really get to a point where you have like a big easy button, you know, you say, give me my SNOP give me my plan. But what we are seeing people do across multiple industries now, including whether it's consumer packaged goods, or automotive, or manufacturing, we are seeing AI used as a facilitator in that SNOP process. Actually one company that we're working with in life sciences. They had 250 custom reports on top of their SAP Implementation to support SNOP. Right? That was before AI came into the picture that was, you know, accumulated over years of using SAP to run their planning. And those 250 custom reports were analyzed by I don't know, call it a dozen supply planners or planners across the globe. And they then used their insights from the manual review of those custom 250 reports to come in with a point of view to influence the SNOP. Now those 250 reports are being loaded into ChatGPT, and we are gaining insights in seconds as opposed to days. And they're still being validated, they're still being reviewed, they're still getting a human approval on, right, the insight. But that was a a massive savings in effort and time with really not much effort at all. Right? You just have to load those reports right up into your enterprise ChatGPT and ask it a couple of questions. And you've got some really interesting insight. Similar insights that those planners eventually would have gotten to. But if you can take, a dozen planners over five days, that's a pretty big cost savings times, what, 12 times a year. So it's out there and it's being used now eventually AI will play even larger roles in, in SNOP. But that's what we're seeing, like, in action today in real boardrooms around the world. And Jen, sustainability tends to be quite horizontal across organisations and across all supply chain decisions, and maybe because of that, it's not often the primary motivator. So in what situations do you see sustainability actually driving the conversation, if any, and and how can leaders make that more common? Yeah. So good question and, And in fact one I personally struggle with as a, a consulting leader. Should there be a a sustainability practice or should it be a capability right that is integrated into planning, into procurement, into operations and I could probably go either way. I keep, I keep consistently going back to the capability and the horizontal because I do think that is how most of the clients that we interact with view it with the exception of those few that have responsibility for reporting around sustainability, right? There's, obviously some portions of the organisation that are dedicated to sustainability. But yeah, you're exactly right. Sustainability is something that is taken into account with virtually every sourcing decision, right? It's, it is one, or multiple questions, right, on, on every evaluation, when you're thinking about acquiring or adding a new vendor to your portfolio in vendor performance reports, in deciding where you're gonna build your new plant and you're looking at you know, transportation costs, and it's a lot easier to have stuff nearshore than and, and actually that's a really timely question with getting back to the, the tariff comment that I made earlier. So yeah, so sustainability gets into every supply chain decision. I think that with AI, we have an opportunity actually, to pull up that sustainability insight because we don't have to kick off another big separate analysis project to give us some insight on what is better or worse for the environment. What is better or worse for sustainability? And so hopefully that gives us some broader application and consideration of sustainability. So I'm not sure if that really answered your, your question, but I and maybe that's indicative of how a lot of companies still struggle with sustainability. Yeah. I mean, obviously companies always look closely at ROI. Yeah. how do you make the financial case for sustainability initiatives when the payback isn't always immediate or straightforward? Yeah, I think in some industries, the, customer pushes that. Right. So I've got two college age daughters, for example, and they're both very into fashion and very into, doing the right thing for the earth. Right. And so for them, it's actually an active high priority consideration where they purchase their clothing from. Whether or not they're doing their Amazon deliveries on a daily, when I want it basis or, having them shipped weekly. There are some consumers in some industries that really are the driving force in that. And I think in other cases, it just makes business sense, right? It is less expensive to use a lighter weight cardboard to ship those packages. And so I think that we should continue to look for those win-wins where there's a a sustainability win and a cost savings or an efficiency win. And they're out there. We just need to make sure that we are putting the right focus on it so that we bother to go looking for those types of answers. And you mentioned tariffs, for example. I think people in supply chain at the moment, particularly in the US or Canada or Mexico or China, I think, they must be feeling a certain amount of whiplash from the way these tariffs come and go with, the likes of regulations and incentives shifting so quickly especially, you know, as the political landscapes change, how do you future proof supply chain strategies so they remain viable, and sustainable under different administrations or policy frameworks? Yeah, so, actually one of the early use cases that we had talked about very simple AI use case was just using AI to look at changing regulations around the world and essentially outsource some of that research to a bot rather than an intern. I I think in under the current US administration it takes a little more focus than a bot, right to respond. I mean, because it, it's kind of like whoever spoke to Trump last, right. Has a you know, has undue influence. And it's not just Mexico and Canada anymore, right. As of now the tariffs on steel globally, right. So it's literally everybody is now caught under this regulatory whiplash to use your apt phrase. So what does that mean that you as an organisation can do to minimise the impact? If you're a manufacturer, you can't be dependent on just China. So your China plus one strategy all of a sudden becomes much more important, but it's not even just that India or Costa Rica can be, a sufficient second choice. I do think that part of what's driving these current waves of tariffs are to push more manufacturing back to the United States. So it's not just multiple sources. It's not just nearshoring. I do think there is an intent to drive it back to American based manufacturing. That will take some time and whether or not everybody buys into that or they just assume that there'll be another news topic of the day next year, or next month, I don't know, but I do think that companies owe their investors, a second and third look around what their global sourcing strategies are whether or not it's for procurement or for actual manufacturing. It probably makes good sense to revisit that anyways, periodically. It's just more front and center right now with all the, all the news, And, when you are talking to companies about AI enabled sustainability, what misconceptions are the biggest? Well, what are the biggest misconceptions you encounter when, when you're talking to them from executives and how do you address those doubts? The number one would be that sustainability costs money and I'm only doing it because it's required. Su sustainability doesn't cost money. Sustainability, you know, if you can think about things in terms of a more holistic approach, sustainability could and should actually save you money and make you more effective and efficient in the long run. I mean, efficiency and effectiveness have long been the foundations of a successful supply chain and that's exactly what drives sustainability. So I don't, I just think that sustainability sometimes has a bad rap of being a PR tactic or just for regulatory requirements. So putting the focus back on efficiency and effectiveness gets us more directly back to making that return on investment argument. I would add resilience to that as well. I I agree. I think that's fair. For people listening who are overwhelmed by AI's potential, you mentioned 76 use cases if I remember correctly, what's one pragmatic first step they can take today to move their supply chain toward a more data-driven sustainable model? So, in terms of use cases, I think using AI to look at some of your existing reporting as I described the early SNOP adopter. That's not a bad easy, low overhead opportunity. I think the more important preparatory activity, though, is to really go back and look at the types of resources that you have in your organisation, and the skills that they have because interacting with generative AI Is different than being an analyst who runs a report. There's a degree of creative thinking of critical analysis. And we were talking about testing the content and that an analyst who generally is collecting data and putting it into an Excel spreadsheet hasn't really been developing or practicing or using a lot in their careers. And those resources are pretty important to most company's supply chains today. But fast forward five or eight years and those skills, pretty replaceable by generative AI or, or even just the old fashioned AI. So I think that companies really need to start thinking about the types of resources they hire, the types of skills that they train for, and it's got to be leaning much more heavily towards creative analytical thinkers. And not so much on rote analysts who are creating reports for the bosses. That's a big transformation that most manufacturing companies are gonna have to struggle with here in the next couple of years. Sure, we've been talking so far about generative AI, we've mostly referred to ChatGPT or Copilot. Have you come across anyone using DeepSeek yet? I know it's it's a bit early, but it has to have massive advantages given that it's open source and it's far less needing of GPUs. So it's entirely possible to set it up in your own IT organisation, train it on your own data and not be giving your data to OpenAI for their ChatGPT models. Yeah, so not DeepSeek in particular, but I think what roiled the market so much right when the news came out was that it opened up everybody's eyes to how much less intensive generative AI could be. I. And how the barriers to entry are actually a little lower than maybe the original generative AI organisations would've let us believe. So, does this open up generative AI to more competitors? Absolutely. Does it make it more open for sure. I don't know that it's gonna be DeepSeek in particular but I think it was just the first to put a chink in, the wall of the current leaders. And it's, getting some of the technologists in generative AI, as opposed to me, like the users and the you know, coming at it from a business angle it shows that there probably will be dozens of different AI players out there as opposed to a handful. Obviously, there's lots of different applications, you know, whether or not you're writing stuff or doing analysis work or doing PowerPoint presentations, right? I, I think one recent report I saw the other day said that the average fortune 1000 company had over 90 different AI tools in their organisation already. So, it's not just ChatGPT and some of the ones that we continually cite. There's lots and lots that are out there already. Yeah, sure, sure. So looking ahead what emerging tech or trend do you believe could be the next real disruptor in supply chain sustainability and and why? Yeah, I, I just think everybody's sourcing strategies are gonna be greatly impacted by the combination of thinking about sustainability from a business case perspective, trying to manage the risks around the next four years of Trump administration and the various tariffs. Obviously it's a hammer in his tool belt that he's going to use. I think sourcing just floats to the top in an immediate application for me of where we can leverage AI to drive better sustainability and resiliency. And as long as companies are keeping the focus on efficiency and effectiveness, it's an easy business case to make. Sure. Sure, sure, sure. Left field question. If you could have any person or character alive or dead, real or fictional as a spokesperson for supply chain sustainability and AI in that, who would it be and why? Oh my goodness. A spokesperson for sustainability ai. My usual go-to answer like anybody you wanna have dinner with right is John Steinbeck. He's my favorite author of, of all time. It's both entertaining to read his books and they're, they really make you think I'm not sure how he would feel about generative AI and sustainability. Although he worked a lot and wrote a lot around farming. He tells a great anecdote in East of Eden about the first shipping of lettuce cross country in rail cars and the ice all melted. And so when the lettuce got there, it was obviously not fit for sale. And that's a, story that that stuck with me in that East of Eden. It was a allegory for some larger themes in the book. So I wonder what he would do with AI and how that would change his lettuce story and advent of some of the early farming and agriculture inventions that he, he wrote about in the 18 hundreds. I dunno that he'd be my spokesperson, but that's what, you know, I, I think he might have some interesting thoughts about it. Yeah. Yeah. You, You just reminded me there of Kurt Vonnegut and I wonder what he would've made of, of this. That would, that'd be an interesting conversation for sure. We are coming towards the end of the podcast now. Is there any question, Jen, that I didn't ask that you wish I did or any aspect of this we haven't touched on that you think it's important for people to be aware of? We didn't talk a lot about what the big enterprise application vendors are doing here in this space. And that is where I see a lot of our clients who aren't naturally early adopters. I keep hearing from them questions or even the comment. I'm gonna wait on SAP. Or I'm gonna wait on Kinaxis, or I'm gonna wait on whatever their application vendor is. That's a, an interesting I mean, it's a perfectly valid approach depending on your, your industry or, you know, your culture and how aggressive or conservative your organisation typically is. So I think keeping an eye on Joule with SAP, we're working closely with them on developing a lot of supply chain specific use cases in Joule. I do think that will help drive a lot of adoption because it will become innate in the applications that they've already deployed, right? It'll just come along with S4 deployment. I thought it was interesting to see Larry Ellison of Oracle standing up at the Stargate announcement in the White House there a couple of weeks back. So it'd be fascinating to see what Oracle come out with as well. But that's a, that's a whole nother podcast, I think. So Jen, if people would like to know more about yourself or any of the things we discussed in the podcast today, where would you have me direct them? Bristlecone.com. You can find me and our whole leadership team there. A lot of interesting thought leadership content around sustainability and AI and obviously supply chain. So, It brings us back to your question about what Bristlecone was named after. So I'll just leave it Bristlecone.com. Great. Jen, that's been fascinating. Thanks a million for coming on the podcast today Thank you very much. I enjoyed it. Okay. Thank you all for tuning into this episode of the Sustainable Supply Chain Podcast with me, Tom Raftery. Each week, thousands of supply chain professionals listen to this show. If you or your organization want to connect with this dedicated audience, consider becoming a sponsor. You can opt for exclusive episode branding where you choose the guests or a personalized 30 second ad roll. It's a unique opportunity to reach industry experts and influencers. For more details, hit me up on Twitter or LinkedIn, or drop me an email to tomraftery at outlook. com. Together, let's shape the future of sustainable supply chains. Thanks. Catch you all next time.