← Supply Chain Now
Loading summary
Transcript75 lines
- [00:00]
Prabhat Panaka
The hardest part of any transformation is never the technology. It's actually getting the people to adopt in a way that genuinely changes how the business operates. Supply chains are full of bureaucracy and because of like age old operations methods. The key lesson which I never forget is that technology has to fit naturally into how people work and not how you wish they work.
- [00:25]
Supply Chain Now Announcer
Welcome to Supply Chain now the number one voice of Supply Chain. Join us as we share critical news, key insights, insights and real supply chain leadership from across the globe. One conversation at a time.
- [00:38]
Scott Lewton
Hey, good morning, good afternoon, good evening wherever you may be. Scott Lewton and special co host Jorge Morales with you here on Supply Chain Now. Welcome to today's show. Jorge, how are you doing today?
- [00:51]
Jorge Morales
I'm great, Scott, thank you, thank you. And welcome everyone.
- [00:57]
Scott Lewton
Welcome, welcome. It's great to see you again. It's good to see you. We were last together in Vegas at Manifest and I know we've rubbed elbows a couple times since then. I really enjoyed your perspective. But Jorge, you set a high standard with your industry perspective and I tell you we're gonna hit it out of the park along those lines with our guests today. Because today folks, we've got a very practical deep dive. It's what I'm gonna call it on AI within global supply chain. Especially from a been there, done that industry perspective. That's a very valuable perspective. We're going to be discussing AI from a variety of angles, including what powers successful scale of AI initiatives, gaps between what I'll call the AI haves and the AI have nots, and how to bridge those gaps. Perspective on critical guideposts that AI initiatives need to have in place. Plus we're going to explore a few of our favorite use cases And Jorge, it always comes back to leadership and people. We're going to be discussing how the golden age of supply chain tech is forcing the evolution of the leadership tool belt and the talent management strategies that are out in the market. All this and much, much more. Jorge, great to have you back with us. Are you ready for today's conversation?
- [02:10]
Jorge Morales
Yes, of course I'm ready and I'm very enthusiastic. I think we're living in very exciting times. Things are changing every month with this AI revolution. Things are moving really, really fast. It seems like it was a few months ago, November 22, when we were introduced to ChatGPT and in 2024, the standards for applications to connect with these AI agents, 2025, the agent to agent protocol. Things are moving at a really, really fast pace and it's exciting to see how things are really happening. And I'm glad that our guest today will speak about how things are happening, what's really happening in the trenches and the real world. Experience what we're talking about here, Jorge.
- [03:12]
Scott Lewton
That's right. Exciting and challenging, but also innovative. And the conversation day to your point is going to offer some wonderful, actionable perspective. So I'm going to dive right in. We're going to introduce our guests first. So our wonderful audience out there, the SCN Global fam kind of has some context. Prabhat Panaka is a product and supply chain transformation leader specializing, as we've been talking about, in AI, digital twins and intelligent enterprise systems. He has led initiatives across fulfillment warehouse operations and enterprise decision support, helping organizations improve resilience, productivity and operational performance at scale. Prabhat's experience includes large transformation programs supporting companies such as Lowe's, Johnson Controls, basf, Dow and Kraft Heinz. He is a frequent speaker and thought leader on AI driven supply chain transformation with a focus on translating emerging technologies into practical operating models and measurable business outcomes. We all need better translators, better Sherpas at all of that. Please join me in welcoming Prabhat Panaka, lead product manager for supply chain at Lowe's as well as an advisory board member at the International Supply Chain Education Alliance. Hey, Prabhat, how you doing?
- [04:35]
Prabhat Panaka
Hey, Scott.
- [04:35]
Jorge Morales
Great.
- [04:36]
Prabhat Panaka
And thanks for a great introduction. I'm glad to be on this one.
- [04:40]
Scott Lewton
Well, great to have you. And Jorge Prabhat has been busy. That is quite a. It's quite a journey we just shared there, huh?
- [04:48]
Jorge Morales
Yes, yes. He's a very experienced professional and we're proud to have him in the ISEA advisory board. He joined us during our panel in Las Vegas where we met in manifest in February. And yeah, it's great to have him here.
- [05:10]
Scott Lewton
Well, it is. And two quick thoughts before we get to the fun warm up. Question number one. I love the element of his background where he translates emerging technologies to those folks that really want to, you know, make it happen for their people and organizations. Blessed are those bridges that bridge that divide. And secondly, folks, you've probably heard me talk about this. I'm a big Lowe's fan. I'm a big Costco fan, but I'm big Lowe's fan. I'll just tell him per bot I'm in the stores. I seems like once a weekend now that spring has hit Georgia, so we'll, we'll probably weave that in somewhere here. But Prabhat, welcome in. We got a lot of work to get to. Prabhat And Jorge, big show. I think it's going to offer a lot of actionable perspective and it'll make people hungry. But I want to do this Prabhat. I want to level set on your professional background prior to your current role. You know, I mentioned some of the companies earlier in your introduction you've worked with. Give us a couple of key roles that really shaped your worldview. Prabhat.
- [06:04]
Prabhat Panaka
Yeah, it's been an interesting journey over the last decade. I started really as like an industrial engineer on the shop floor doing time studies, kind of hunting for process inefficiencies. That's where I kind of like, you know, developed a deep appreciation for what technology can do because often not like, you know, removing waste kind of like meant like you had to implement some sort of like technology system. And so when it is thoughtfully applied and I think technology can be a game changer is the first lesson that I learned from my industrial Internet days. And then further on I kind of pivoted into a purely supply chain strategy and technology consultant with a couple of big four consulting companies working on large scale transformations for Fortune 50 companies in supply chain. Right. And the hardest part of any transformation is never the technology. It's actually getting the people to adopt in a way that genuinely changes how the business operates. For example, I had a client, a large chemicals company go through like about 6 advanced planning system implementations and each one struggled. And the technology was good on paper and was good when implemented, but people never fully embraced it. And as you probably know, supply chains are full of bureaucracy and because of age old operations methods. And that's what was the key lesson, which I never forget, is that technology has to fit naturally into how people work and not how you wish they worked. And that kind of experience fundamentally shapes how I think about product design and supply chain systems today in my current role too.
- [07:54]
Scott Lewton
You know, Jorge was smiling during a portion of your response and I'm sure it's because he's heard that we've all heard that challenge time and time again. It oftentimes and there's plenty of technologies that, that you know, aren't great, but oftentimes there are great technological solutions. And the greatest challenge is one of, one of those that you, you called out and that's getting our hard working talented people to adopt the new technology, the new process, the new chapter that transformations bring. Jorge, I know you've heard this quite a few times before, huh?
- [08:28]
Jorge Morales
Yes, definitely. A lot of people think that we need to be in the case of AI, which is you can apply that to any technology. But AI is something we're now all already familiar with. We are now all AI aware. So we kind of understand what it is. But a lot of people think we have to be AI proficient. And a lot of people dedicate their whole lives to becoming AI proficient. So that's a skill that's difficult to get. However, in the middle, there's the sweet spot in which supply chain professionals should be, which is they need to be AI competent. And what I mean by that is they need to be. Be able to harness the technology, understand the language, understand what technology can do, what it can't do, and be able to. You don't need to understand. We made that analogy some time ago. But you need to understand what's under the hood of a truck in order to use trucks to move what your business is producing. Right? So you can harness the power of trucks. And the same thing happens with technology. But you, you need to understand how to use the trucks, what's around all the trucks, what, even the details when you get into the curve and there's a bank and need, what speed you get, you need to. But so those are the kind of things that you need to understand where the limits between being proficient and aware and what you need to know to be competent and the standards, what are the guardrails, what's the knowledge that's really needed. In isea, we focus a lot on that. We focus a lot on getting our certification programs, csetp, the Technology Professional program, focused on that. And now we're getting a spin off from that, the Certified Professional in supply chain AI. So that's coming out very, very soon because things are moving at a really fast pace and people need to understand that sweet spot. They need to understand what they need to know how technology can be used. So I love that Pravat is being, working with all that. And we're very proud to have people like Prabhat in our advisory board and helping us get our programs better and better every time.
- [11:14]
Scott Lewton
I'm with you. All right, We've talked about the need to get at least AI competent. You don't have to become an AI subject matter expert. But I'm going to try to get my AI competent certification from Prabhat in this conversation. Jorge's already got his. But I think also for important context, before we get into the AI part of the conversation today, Prabhat, if you would tell us more about your current role and what you do at one of my favorite companies being Lowe's, and then also I want to touch on what you're doing with isca, but tell us what you're doing at Lowe's first.
- [11:43]
Prabhat Panaka
Yeah, for sure. And thanks for being a great customer. And we want to make our stores as seamless for like, you know, homeowners like yourself to come and shop at it and make it a memorable experience. And so thank you. So in my current role as lead product manager for supply chain tech at Lowe's, what I do is like I focus on building the core systems that power our warehouse operations, which supports our 1700 plus stores across continental U.S. primarily it is going to be our warehousing management systems, which I call it as like the system of record for everything that moves through our facilities. Right. So you might. So this work is kind of like very foundational because here's what I believe. You cannot layer AI or agentic AI on top of like fragile infrastructure or lack of data or lack of process and kind of expect it to perform. So the data has to be clean, the systems have to be well integrated, and the kind of logic that powers them has to be sound before you can trust AI to do some sort of actions on your behalf and not just advice. Right. And that's exactly what we are building a future where AI isn't surfacing recommendations, but kind of handling exceptions, triggering workflows and escalating with the right context, which I call as bounded autonomy, which means the AI that operates within defined rules and knows when to escalate to humans. I think that is going to be the key for tomorrow across supply chain. So the work that I'm doing is laying that foundation and that makes AI future safe, effective and coverable in terms of execution in tomorrow's supply chain environment at Lowe's Prabhat.
- [13:33]
Scott Lewton
Love it. And when you say tomorrow, you're not talking about Thursday. You're talking in a greater sense in terms of the next age we're moving into.
- [13:40]
Prabhat Panaka
Yeah, next thing is never moving too. So that's a good call out.
- [13:44]
Scott Lewton
Hey, and really quick, I tell you, it's like supply chain utopia. You paint this picture. Clean data, integrated systems, sound logic. Man, when you've got those three things, fundamentally you can move mountains, not just the supply chain, but elsewhere. Quick follow up. To do all of that stuff and really operate a successful organization here in 2026, you got to continuously learn new things, right? And of course we reference your role as a board member at the the one and only is C E A which Jorge leads as global coo. Tell us about importance of continuous learning
- [14:22]
Prabhat Panaka
for Bot, I think continuous learning is a principle that every industrial engineer or every supply chain practitioner kind of carries with themselves right from that core principles of like lean thinking. But I would say that like as kind of AI becomes central, right? The role itself is changing and supply chain practices are not kind of isolated from it. So what I mean by that is that like your operations or supply chain operations is focused on doing what I call as hands on coordination task. And that would change into what I call as like, you know, system architect roles. You know, people who design AI systems, govern AI systems and then coach the surrounding teams on how to like, you know, manage the shift. So like, your role as a supply chain operator in my opinion is going to shift from like, you know, that coordinator to more of a systems architect and a coach. And what does that mean from a learning perspective? You have to be prepared for that evolution of the job role. And that's where I think organizations such as ISEA play a very crucial role. They provide the certifications and the trainings on the topic, which are infused from latest industry insights from people like me and others on the board. And I also selfishly believe that being part of ysea, I'm like kind of like sharpening my own skills by kind of hearing insights from others on the board and then simultaneously helping Jorge and his team prepare the best learning material that could be there for supply chain practitioners so that we can all evolve together in the age of AI enabled supply chains, which is not going to be very far. I think by 2013 you're going to have like, you know, major enterprises working a large part of it, like you know, working on autonomous supply chains, man,
- [16:22]
Scott Lewton
triple win is what I'm hearing there. And Jorge, going back to the continuous, the value of continuous learning, which is almost an imperative. When you and I were talking and I interviewed you in Vegas, you kind of put it bluntly and I pulled up the blog article here to my right. Adaptation or be left behind? Well, the only way to really adapt is be able to train and what you must learn and train on what you must adapt to. Your thoughts, Jorge?
- [16:48]
Jorge Morales
Yes, as I said, these are exciting times. But you can either do the homework, roll up your sleeves and get into it, you can make that decision in try to remain current, to keep up to date and keep your job,
- [17:13]
Prabhat Panaka
or
- [17:13]
Jorge Morales
you can risk being replaced by either an AI agent or someone who can harness the technology in a better way. So I think it's very, very important that we become self conscious to understand more. Not all of us are familiar with Code programming, but we can use them and we can learn to use them very effectively. By developing this network of agents and using harnessing these tools, we can make our companies more efficient and make ourselves more valuable for the company.
- [17:57]
Scott Lewton
Yes. So folks, if you don't get your continuous and regular learning opportunities from iscea, get that help somewhere. It's imperative because as Jorge said, you may not be replaced by AI, but you will definitely be replaced by someone that uses AI better than you do. Right. That's a big threat. Not for everybody perhaps, but kind of depends on your background, what you do. Let's dive in, provide. I want to get into AI because there's a lot of, for all the successes out there, there's still a lot of heartburn and friction from those organizations trying to find their own. Right. AI gear. So I think you've said in the past the conversation is moving thankfully beyond pilots. So what separates its organizations that actually scale AI from those stuck in endless experimentation? Prabhat.
- [18:49]
Prabhat Panaka
A lot of my colleagues think that technology is the differentiator. But like, technology is not the real differentiator. Like, it's the operating model around how you enable that technology is the differentiator. So what I've seen is that companies that get stuck in pilots, it's about lack of structure and accountability. In my experience, successful organizations do is they try to embed AI into workflows and with clear defined ownership. And they're fine with starting with good enough data. And it's a common theme that data garbage in and garbage out, right? So when it comes to data, if you end up waiting for that data to be correct, you will not move beyond that pilot to production. So they take that calculated risk in moving forward with that good enough data. And one of the other primary things that I have seen where organizations have become successful in moving AI pilots to productions is setting out like, you know, what objectives that they're going to chase with this AI decisioning tool or AI workflows. Right. Either it can be cost of doing some activity or the speed of it. Like how do you measure it effectively and how do you make sure that it's actually moving the needle? And this all kind of informs us to a point where that model is actually successful. And it's kind of like giving the necessary outcomes that we had designed it to be. And when you want to move it to production. The real challenge which I kind of touched upon early on is like change management. Right. The difference between AI and some automation capabilities. AI is more of a probabilistic Output and automation is more deterministic in nature. So there's a little bit of learning that we need to do as operators in order to learn how to digest that data point and make decisions upon that data point, trust that data point. So ultimately people need to trust and use it. What I've seen is that the pilot is successful, it takes on all of those goals, but then it fails when it moves to production because people are not able to make sense of the data that is being recommended out of the AI model. And finally one more thing, right? Like we need leaders who are willing to accept a learning curve. And this technology is kind of new. It's being implemented in companies right now. So there's a lot of like unknowns that are there, right? Like you should be comfortable with like you know, the first bump, let's say you decide to launch it, but it doesn't go well. You should be comfortable with that. And so in summary, I would say like, you know, making AI successful from pilots to production is all about structure, is about ownership, is about defining success from the start and being uncomfortable and going on that learning journey.
- [21:57]
Scott Lewton
Prabhat, love it. Now what I heard, I loved your summary too. I'm going to add my own. This is what I heard from Prabhat Jorge. The data, the targeting of the right problems with AI change management, of course the outcomes and the data outputs have to be actionable and be trusted. And then of course every what is not about leadership. You know, leadership is like I learned in science in second grade that water is a universal solvent. Leadership is also a universal solvent and of course has a place when it comes to AI. Jorge, what'd you hear there from Prabhat?
- [22:30]
Jorge Morales
Yes, I agree with provide 100% on measuring how good are you doing and getting the results. Because based on that you have the arguments of moving forward to going further with these initiatives. And I think what Prabhat said also makes sense with what you said about leadership. If the leadership is aware that there are going to be bumps and things are not going to be working as expected from day one, things are going to be easier. It's going to be like Napoleon used to say, dress me slowly because I'm in a hurry. So if you want to, if you hurry the when you're dressing up, you might do it wrong.
- [23:32]
Scott Lewton
That's right. It's interesting to call out based on what both of you are sharing here. Even with the perfect structure and approach, there's going to be bumps and there's going to be friction. All of Them are reason to really do your homework on the front end to driving change with AI using whatever tool is finely tuned. All right, so let's talk about this prabhat. The biggest gap today that you see between what AI can do in supply chain. Right. The art of the possible, or what I'll call maybe the art of the practical possible. Right. And what companies are actually implementing and doing with it. What's important to note about that gap?
- [24:11]
Prabhat Panaka
Yeah, I wrote a recent article about it and it kind of resonated with like multiple supply chain leaders on LinkedIn. But the biggest gap that I kind of talk in that article is like, you know, most companies still treat AI as kind of a recommendation tool when it's kind of fully capable of being an ex execution engine. So here's what I mean. Like, you know, today companies use AI models to make surface decisions. For example, demand forecasting. You need to like, you know, order a certain product. And then with the coming of age of like ChatGPT and LLMs and generative AI, they put a chatbot on top of that, like recommendation. And now you can query in natural language. And that's the most that like most companies have done. But the true power that I talk about in that article is about what I call as compressing the detect, decide and act loop. So for example, take this. You have a warehouse, warehouse operations damage pallets often show up at the warehouse today. What happens is that typically that gets put aside and it reached for a human to do something about it. Imagine tomorrow like AI embedded in that workflow. Like, you know, agentic AI could immediately place a hold, open a supplier claim, notify the right people, all within the right context, all without doing any warning. Right. And that is where I think the real value comes out of is like when you embed AI into those execution workflows. So the technology to do that is existing today. So I think it is just an organizational willingness to let AI execute and not just advice. And I think the core challenge in that is defining what I call as bonded autonomy. So that is being explicit where AI can act independently and where it needs to escalate to a human. So the whole human in the loop concept and that boundary is not a technical problem. I think it's a leadership decision issue. So until that companies can make that shift from insight to execution, they're only capturing a fraction of what AI can deliver. So move from beyond chat boss to execution. Embed AI into execution is going to be my biggest advice.
- [26:43]
Scott Lewton
I like it. Prabot. Jorge, you hear that back to your pickup truck. AI can be the execution engine in that pickup truck you're using earlier. What, what did you, what did you gather from what Prabhat just shared about how to in a way cross the gap, cross the chasm and really realize and leverage more of what it can be, rather than just that recommendation tool that like 95% of humanity is using it for. Let it be that execution engine. Your thoughts, Jorge?
- [27:16]
Jorge Morales
Yes, I think AI agents have the great potential of autonomy and that opens a lot of opportunities and possibilities. However, I still think that we need to be careful. And that's part of what I was saying about being AI competent. You need to know what AI can and cannot do, but you also need to understand how and what level can you rely on AI. And you still need to understand the issues that must be handled from the cybersecurity perspective, for example. That's the reason why it's very important to learn more.
- [28:05]
Scott Lewton
So Jorge, you're reading my mind where I won't go next with Prabhat, because one of the things you're kind of implying and speaking to there, and you mentioned this earlier in one of your responses, is guardrails, right. Before we can broadly trust and really allow, give permission, give human permission to AI to make or execute those decisions truly autonomously. Because that's what is the practical possibilities today, Prabhat, what guardrails? When it comes to data or governance or organizational, what must we got to put in place, Prabhat?
- [28:37]
Prabhat Panaka
I would broadly organize it into three layers. The first being around data integrity. About this, autonomous AI decisions are kind of like only as trustworthy as the data feeding them. Like bad data in, bad data out. Right?
- [28:52]
Scott Lewton
Right.
- [28:53]
Prabhat Panaka
So before you let a system execute autonomously, not just recommend, you need to know where the data comes from first, how fresh is it? Because there's a concept of like data drift and that leads to a decision drift and you need to know what happens when it's wrong. So kind of like have a sail fail mechanism. So most organizations haven't answered those questions honestly. So that is like the first layer of execution or operational capability that you need to develop is understand data integrity. Right. And the second, according to me is setting up these decision boundaries. You have to be very explicit about like what AI is allowed to decide versus what requires a human in the loop. Think about like, you know, exception based planning and supply chains. Right? It's similar. You have to do like, you know, exception based autonomous decision making. Right. And I'll give you like one example. Think About a system that can automatically reorder $50,000 of inventory is very different from the one that can cancel a $5 million supplier contract. So there's magnitude of impact that you need to like take into consideration when setting up these decision guardrails. And these boundaries need to be designed deliberately and not kind of discovered after something goes wrong. And that is where I think it's very important for tomorrow's supply chain operators to do. Understand how AI operates and understand what AI can do so that you are enabling these decision boundaries. And the third is what I think is about auditability and traceability. What I mean by that is that when a autonomous decision taken by AI kind of causes the problem and it is going to bound to happen at some point, you need to be in a position to explain why exactly the system did what it did. And it's not for compliance, but actually it's because you can get the feedback to improve the system and maintain that organizational trust in it. It goes back to my key point that I mentioned earlier is like any technology is as good as like the organization that adopts it. If you lose the trust, then the organizational adoption goes away. And if you can't explain that decision, you're not ready to automate it. Really. So those are the three things. Data integrity, decision boundaries and the ability to audit.
- [31:34]
Scott Lewton
Yes. And if you don't have answers to those questions, you're going to get in trouble. Per bot. Also, hey, but really quick when, when you talk about those decision boundaries, Jorge, and he was using that kind of example. That was a good example, right? The 50,000 versus the 5 million. But also kind of remind reminds me especially when we think about, you know, how a lot of folks use that like child analogy with AI as it continues to kind of mature and get older and older. My brain goes, Jorge, to my three kids and the decision boundaries I have in place with them for their allowances or for any, anything. Yes, you can buy that five dollar cup of coffee, but no, you could not buy that $500 dress or Amazon order. Jorge, is that you have those similar decision boundaries.
- [32:22]
Jorge Morales
Yeah, as you know, I have four kids. And yeah, the same thing happens. And I think companies are facing the same issues because as Pravad said, AI agents are like having another member in the team if the rules are not in place, if those, if those guardrails are not preventing it from doing damage. So it's yes, the same thing happened with my kids. And of course they evolve, they learn they're not the same guardrails they're for my 9 year old than for those who are already graduated and making the wrong rules. The same thing will will happen with eventually.
- [33:19]
Scott Lewton
And you know, I love how practical you kind of captured those three guardrails when the guard wells and think of governance, there's all kinds of insecurity, all that, there's all kinds of usuals. But I really liked your how you tackled those three exams. There's a long list of guardrails, but those three data integrity, decision boundaries and of course that very valuable audibility and traceability if I capture that. Right. All right, so Prabhat, when it comes all together, right? Because a lot of your journey you've been using all a lot of what you've been sharing here and you've been driving outcomes. So when you think of one of your favorite recent use cases associated with the implementation of AI Prabhat, what's one of your favorite stories here recently?
- [34:05]
Prabhat Panaka
Yeah, this is one interesting example and this comes from a peer study that I did with some of ISEA board members and how agentic AI kind of is changing the role of supply chain operators. So this example comes from a membership based wholesale retailer, think high volume distribution centers processing enormous amounts of inventory daily. Right. And what they did is that they deployed a computer vision AI to monitor inventory flow in real time, which basically meant automatically detecting defects, damages and kind of like labeling errors as product moved through the facility. This goes back to the example that I gave earlier in the podcast. So what according to me is remarkable was not just the application of the technology, but it was the operational outcome. So these claims used to take days to resolve. It used to wait for somebody to act upon it, for somebody to like adjust the inventory issue, supplier claims, notify the stakeholders. Now all of it was being handled by this AI agent autonomously in minutes. So your inspection cycles compress from like days to minutes. And this goes back to the whole concept of compressing, detect, decide and act loop. Right. And here's the other interesting aspect. This is the most interesting aspect to me. The manager's job description essentially wrote itself, right? The DC manager went from personally executing these operational tasks to now governing the edge cases that AI couldn't handle. What are the exceptions? That AI needs a human in the loop. So he was just managing now and his work went up in scale of what I call value chain kind of managing vendor relationships at a systematic level rather than at a transactional level is one example. And he gave me this statement, he said like my role shifted from operational execution to strategic Governance. And I think that's a real use case of like AI, not just saving time, but fundamentally changing what a supply chain leader will be doing tomorrow, not,
- [36:29]
Scott Lewton
not Thursday, in the bigger picture. Next chapter. Sorry, I keep coming back to that. Prabhat. I love it. All right, so Jorge, I love that example and here's why I'm trying to keep up notes fast and furiously. It's tough to keep up with, Prabhat. You already know this, Jorge. The operational, the sheer operational outcomes, right. The efficiencies gained compressing that cycle, right. All the days, the manual work, getting rid of a lot of that. You're going to be delighting customers and probably some suppliers based on the supplier's role in the inventory decisions we made. And the best part, perhaps Prabhat and Jorge, which you finished on, which is really elevating the value and the role of the humans as part of this inventory initiative. That's my favorite part. Jorge, your thoughts?
- [37:21]
Jorge Morales
We have wonderful opportunities ahead of improving our processes from a different perspective. Not just the efficiency or making things faster or these repetitive tasks improving quality, all those will be handled in some way by these AI agents. But there many ways in which our processes will be improved strategically in terms of innovation, in terms of having these minds becoming more creative, becoming, dealing with what AI is not capable of dealing with like procedural competence or creativity or coming up with completely new things. Because AI learns very well from the past. But it is our job to envision where we're going next, what things have not happened yet and how things can improve in a different way or finding a new direction. So those are the kind of decisions that will be made by humans and supply chain professionals capable to understand that and understand where are they adding value to, to, to the company, to the processes. I think that's, that's the place where we all want to be.
- [39:03]
Scott Lewton
So let's do this, Prabhat. Yes, we're going to put your, your advisor hat on here, Prabhat. If you're advising a global enterprise that we're starting their AI journey today, they're getting late. What a late start. But that's what, that's what this, that's what they're doing. If they want to drive real impact within six months, not six years, what is your golden piece of advice or two that you challenge them with?
- [39:30]
Prabhat Panaka
I think there will be three things that I would tell them immediately. The first is to start with a decision, not like, linger upon, like a data set. So I think most companies need to start by asking what data Most companies do start by asking what data do they have? I think the better question is to ask what decisions do we make repeatedly that are slow, expensive and inconsistent. Find those kind of work backward to have the data and the models you need. And that framing alone, like, you know, will save you 18 months. So you, instead of starting chasing data, you should start chasing like identifying the decisions and the problems that you need to solve. Right. And the second, I think is the most important one is to find a business owner, not a tech owner for every AI initiative. So I've seen this like multiple times. The only person that is accountable for deploying AI is the IT or data science team. It will never actually scale because the scaling part comes as like, you know, moving from pilot to production. Right? And that's where building trust is very important. And you need somebody's business performance directly attached to that, like, you know what that AI is going to do and that is what is going to move the needle when that AI system is going to be in production. And the third is, I touched upon this earlier, that you should be comfortable with your first deployment being a very, very humbling experience. And what I've seen over years is the companies that scale AI are not the ones like, you know, who got it right first. Even think about like Google. Google was much more further than ChatGPT, but they still waited, waited, waited. And ChatGPT got launched by OpenAI. Right? So it's not necessarily that you get everything right the first time, but the ones that are successful become successful because they build the organizational structure and reflexes to learn fast, adapt and kind of keep experimenting. And so in essence, resiliency matters more than perfection in early stages. Sticking to the strategy of deploying AI. And so this was like one learning that I have had over years is that enterprises that are good at doing technology aren't necessarily the ones better with strategy, but they started earlier and they kind of stayed consistent along that journey. And that is what is giving them benefits when like, you know, it comes and marries up with business outcomes.
- [42:12]
Scott Lewton
So Prabhat, that was really good, those three pieces of advice. It was so good that you might be tempted to send us an invoice and don't do that. Okay? I can't, I can't afford your invoices per bot. But kidding aside, Jorge, that was really good. Very practical, been there, done that advice for companies just now kind of getting their feet wet, wet and standing up their AI initiatives. What you think, Jorge?
- [42:36]
Jorge Morales
Yeah, I couldn't agree more. It's very important to understand the problem first, what needs to be solved instead of what can I do. The way Prabhat put it, I think it was amazing because sometimes you know what you can do because you have the data or you have the tools or you have access to some other third party data. But what you can do might not be the thing you need to do. You need to start deciding first what's that need that needs to be addressed through technology. And that's the main point. So I totally agree with Prabhat and
- [43:26]
Scott Lewton
starting with the first one, starting with those decisions, especially if those decisions can be stressful, regularly interact inaccurate and multitudes of decisions taking tons and tons of time. So I like that. All right, so for the sake of time, we're going to bring it back to leadership. I was going to talk, ask you about talent, but I'm going to ask you about leadership instead. And then we're going to make sure folks know how to connect with you, Prabot, and you Jorge. And Jorge, you're not getting out of giving me your favorite key takeaway from what all the brilliance that provide is, is sharing with us here today. So get ready for that. Your key takeaway provide. That's what we said earlier. Everything comes back to leadership. Right. One of the big common themes you, you've said here about leadership throughout all of your perspective is how we've got to get comfortable with the bumps and the dead ends and the, the failures as we try to, as we experiment and we try to craft the right strategy that brings the desired outcomes. That doesn't come natural for a bunch of folks or a bunch of leaders out there. So along those lines, how does the optimal leadership tool belt change as we have long since entered the golden age of supply chain technology?
- [44:42]
Prabhat Panaka
Yeah, I think what I've come to understand is that the leadership tool belt changes in a pretty fundamental way. What I mean by that is that in past or in the current environment, a lot of like good supply chain leaders do great is about coordination. Right. It's about like following across functions, managing exceptions and like making sure that kind of information moved from one team to another. Right. So it's about like you know, driving that operations day in and day out. Right. So in the future when technology is going to start to take about more of that analytical and coordination work through autonomous agents or robotics. So I believe the leaders role kind of like shifts upward in the value chain. Right. So what I mean by that is that this is a concept that I kind of alluded earlier on to is leaders Increasingly have something like system architects and coaches and system architects in the sense they need to design how workflows, data automation and decisions fit together and coaches in the sense that they need to help teams learn, adapt and trust in the systems right. And know when that human judgment still matters. So I would say the 2030 Model Leadership Tool belt and supply chain is going to be less about supervision, more about systems thinking, change management, exception judgment is one more key area and then guiding human AI collaboration that to me is a big shift. I think the work that Jorge is doing with ISEA kind of like helps supply chain professionals navigate that change. Over the next few years Prabhat we're
- [46:34]
Scott Lewton
going to need a lot more technology sherpas and and trusted advisers and a lot more Prabhat Panakas as we navigate the tons of opportunity but also tons of change and tons of who moved my cheese like that old book tons of of new ways of doing things. It's all exciting but it al also creates some consternation for a lot of folks out there. Jorge respond to how Prabhat talked about the evolving optimal leadership tool belt.
- [47:08]
Jorge Morales
I think it's very important that all of us take our place or embrace the responsibility that understand what's our part in in in all of this. Because yes leaders need to to define the direction leaders need to motivate their teams. They need to become pro technology. But ourselves as individuals, we all, we also need to be ready And I think that's that's something very important and a key takeaway from what Prabhat said today is make that decision but that decision not just for the company, not just as a company leader of deciding to implement or move in the AI direction but yourselves. Every one of us need to make a decision on becoming AI competent, on learning more, on getting into that sweet spot in which all these technology jargon is not like gibberish or words hard to understand that we can understand what technology people are talking about with not to be proficient. We don't have to be that subject matter expert but we need to understand how to harness technology and I think that's very important. Learning, becoming certified and also attending events. Next October we have our supply chain technology conference. I'm inviting all of you to attend either virtually or in person. Now it's going to take place in Malaysia because that's an itinerant event but you can attend virtually and learn from all these experts. Prabhat was there last year and if you register you can watch his recording because we're making all the previous year. Recordings available to those who register for the virtual program and learn more, get more perspectives, get more knowledge and make that decision to change your mindset into becoming AI competent.
- [49:47]
Scott Lewton
I love it. And folks can go. How can they find the SC supply Chain tech what site Jorge, for folks to find information on that?
- [49:55]
Jorge Morales
Yeah, it's sctechshow.com you can find more information about that particular event there. But also on the ISEA global website, isea.org and you can find more information about the ISEA certification programs there, the technology certification programs and other certification programs for supply chain professionals.
- [50:24]
Scott Lewton
Outstanding.
- [50:25]
Prabhat Panaka
I'm also actually speaking at 20261 2. Hooray about that.
- [50:30]
Scott Lewton
Okay, late breaking news right here on Supply Chain now. Late breaking news per bot.
- [50:34]
Jorge Morales
Yeah, we have a great lineup of, of experts and that we have. Thank you. Thank you Prabhat.
- [50:42]
Prabhat Panaka
And the topics might be interesting to the audience. It's about like you know, talking about ontology as an enabling layer for agentic AI workflows.
- [50:54]
Scott Lewton
Okay.
- [50:55]
Prabhat Panaka
So ontology kind of refers to like you know, stitching upon like all of your data together in a knowledge based graph so you can put an agent on top of it and then like have the agent kind of like navigate these decision points across cross domains. So it is an interesting topic.
- [51:14]
Scott Lewton
It is call it out sctechshow.com for that fall event and then also in the broader sense you can go to is c e a.org Is that right, Jorge?
- [51:28]
Jorge Morales
That is correct.
- [51:29]
Scott Lewton
Okay, all right, so let's. Did you see how sneaky Jorge was? He worked in those key takeaways in that same question. So and there's key takeaways get ready but not just get ready, got lean in and engage. And not just because if you're a leader, like a formal leader, because we're all leaders, right? We really are if you choose to be one. We've got to really put our own fate and our own success in our own hands. And we all have an opportunity to do that each and every day. So don't let, don't wait till your organization says hey, do you want to go do this look for these learning opportunities now. Okay, so Prabhat, other than your esteemed keynotes that sounds like you've got coming up, how can folks connect with you? Prabhat Panaka?
- [52:16]
Prabhat Panaka
Yeah, you can find me on LinkedIn Prabhat Pinaka. And you know, connect with me there. And I would be interested in learning about like others problems, the ways that they are doing different from like what I just described it's always great to connect with folks from like supply chain industry. It's kind of like a very small world. That's what I realized that Manifest. So I think a lot of familiar faces and names might be watching this podcast.
- [52:46]
Scott Lewton
Maybe so. Maybe so. You know, it's a, it's a big, small world, Prabhat. It's kind of what I've, I've come to realize about global supply chain. A lot of folks know each other, but you're, you're constantly meeting new companies and new, new startups and new leaders. You know, I love. The one element about supply chain I really love is it brings people all the time from other sectors and other functional areas of global business. And that really is a strength of our industry, I believe. Jorge, how can folks track you down?
- [53:20]
Jorge Morales
You can connect with me through LinkedIn and also you can write me directly to my email. It's Jorge or with J OR George M. ISEA.com so you can connect with me. And also we can all go to Javier's in Vegas because we're having our ISEA prize ceremony in Vegas next February. So if you're going to be there for Manifest, be sure to join the ISCA recognition ceremony. And after that we can go to Javier's to have some Mexican food.
- [54:08]
Scott Lewton
Let's do it. Let's do it. So folks, you can learn more about all of that@isa.org you can also email Jorge. You can also connect with both Prabhat and Jorge on LinkedIn and they welcome your feedback and they welcome your what you're doing in your neck of the woods with AI, with supply chain or learning, you name it. Of course there's a slew upcoming events that Jorge and the team are hosting, so we encourage you to get connected. Okay. What a great conversation. Wide ranging. I think I did get AI competent as Jorge called it, right. AI competent earlier. So thank you, Prabhat, for that. And if you haven't gotten AI competent yet, Jorge, they can give you a call. You can swing by on your pickup truck and you can take them to ISEA training so they can get competent. Is that, is that a good Jorge?
- [54:58]
Jorge Morales
Yeah, sure. I'd love to have company.
- [55:03]
Scott Lewton
All right, all right. What a great conversation, folks. Hope you enjoyed it. Hope you enjoyed as much as I did. One big thanks to Prabhat Panaka, of course, with Lowe's and with isea. Prabhat, thanks for being here, my friend.
- [55:15]
Prabhat Panaka
Yeah, the pleasure is mine. And I would close this by saying that tech fluency is a must and Every professional in every industry need to be like tech fluent. And the reason why I say that is ChatGPT has made it so easy to use technology, interact with AI and they are like, you know, close to a billion users. So that means like, you know, it is going to be a part of your workflows tomorrow. So yes, along that journey of tech fluency. And it's a great show, Scott. I enjoyed being here with you and Jorge and looking forward to like, you know, the next episode, whenever that would happen.
- [55:57]
Scott Lewton
It'll happen soon. I appreciate all of your kind words. It's a great episode, great conversation. Also, big thanks to my esteemed co host, Jorge Morales. Always a pleasure, Jorge, to connect with you.
- [56:09]
Jorge Morales
Thank you. Thank you, Scott. It's a pleasure. Thank you.
- [56:12]
Scott Lewton
I appreciate what both you are doing in industry to make us all better and unlock all sorts of innovation and good change ahead. Okay. And by the way, Prabhat mentioned is challenging, folks. You know, we've had all kinds of platforms make it easier, but one thing AI is not going to do is drag you out of bed, drag you into learning and workshops and force you into, right, the opportunities we still as humans have to make that decision. It's just how it works. At least for now. We'll see what next. We'll see what tomorrow brings. Prabhat. All right, so to our SCN Global fam. Hope you enjoyed the conversation, but you know the homework, right? Prabhat and Jorge both brought some very practical, actionable perspective. Take one thing from what Prabhat or Jorge shared here today, do something with it, right? Share it with your team. Put it in action. Deeds, not words. That's how we're going to keep transforming global supply chain and leave no one behind. And with that said, on behalf the whole team here, Scott Luton, challenge you do good, get forward, be the change that's needed. We'll see you next time right back here on Supply Chain Now. Thanks for buy
- [57:16]
Supply Chain Now Announcer
Join the Supply Chain now community. For more supply chain perspectives, news and innovation, check out supply chain now.com subscribe to Supply Chain now on YouTube and follow and listen to Supply Chain Now. Wherever you get your podcasts.