Nick Gospodinov (00:00):
If you think of a traditional software service systems, it’s kind of like we were promised this system that does work for us, but since we’ve had this technology, we’ve actually had to hire specialist people that work with this particular ERP, this particular procurement system, whatever kind of software specialist, because it takes so much time to actually work to the point where I actually hire people to work for the software instead of having software work for me.
Voice Over (00:29):
Welcome to Supply Chain Now, the number one voice of supply chain. Join us as we share critical news, key insights, and real supply chain leadership from across the globe. One conversation at a time.
Scott W. Luton (00:42):
Hey, good morning, good afternoon, good evening, wherever you may be. Scott W. Luton and the one and only Rick McDonald with you here on Supply Chain Now. Welcome to today’s show. Rick, how you doing today, my friend?
Rick McDonald (00:52):
Scott, I’m doing great and it’s a tremendous honor to be back with you today and I’m looking forward to our show.
Scott W. Luton (00:58):
Man, I am too. And I’ll tell you, we’ve had some great conversations over the years and I think we’re going to be able to add this one to that inventory. So folks, some data suggests that global supply chains move more than $26 trillion in goods each and every year. Yet behind all that mind-numbing complexity and technology, a remarkable amount of work still gets coordinated through emails, phone calls, spreadsheets, who knows, maybe fax machines and the worst. And don’t be surprised, let’s not act like we’re surprised by that. So what happens when AI starts connecting these dots, finding the latest truth and taking action before humans even have to chase it down, maybe before they even get up in the morning? Today we’re joined by an innovative leader to explore the next frontier of supply chain coordination, the role of AI agents, and what it might actually mean for supply chains to one day coordinate themselves.
(01:51):
That’s pretty exciting. And Rick, many members in our audience will recall that you served as chief supply chain officer with the Clorox company and given your impressive supply chain leadership track record, including all sorts of transformation and innovation outcomes, I can’t wait to hear your take on today’s conversation. Should be a good one, huh?
Rick McDonald (02:10):
Yeah, this is really exciting space, Scott. It is one of the areas where I think there’s the most opportunity to advance the capabilities, whether they’re just artificial intelligence or agents in some way expanding and advancing in the procurement space. I think it’s going to be a great conversation.
Scott W. Luton (02:27):
I agree and I can’t wait to dive in. So folks stick around for a great discussion that’s going to offer up tons of actionable insights by the container load. So I want to welcome in a special guest joining me and Rick here today. Nick Gospodinov is the co-founder and CEO of Mandel AI. Now he’s an entrepreneur focused and passionate about solving complex supply chain challenges, particularly how manufacturers and suppliers communicate, coordinate, and respond when things change. Let’s welcome in Nick Gospodinov with Mandel AI. Nick, how you doing my friend?
Nick Gospodinov (03:05):
I’m doing amazing. Thank you for the warm welcome, Scott.
Scott W. Luton (03:09):
Hey, you bet. You’ve been on our radar for quite some time. I tell you y’all have been on the move and I can’t wait to dive more into what you’re doing and why it’s consequential for our industry. And of course, we got Rick McDonald here as well to weigh in. So let’s do this. Nick, I want to start with a little fun warmup question. And I learned, I think Rick and I both learned in the pre-show one of, I don’t know, one of 117 reasons why you don’t want to mess with Nick. And that is because aside from the supply chain prowess, you’re all about one of the things you do on the side, I guess as a hobby, martial arts and all sorts of different martial arts. Tell us about this, Nick.
Nick Gospodinov (03:51):
Yeah, absolutely. So I got into martial arts, I think it was just right after COVID. So I’ve done boxing, I’ve done Muay Thai, I’ve done Brazilian jiu-jitsu. And I think it’s one of these things where on the outside, it looks like it’s a very violent art, but I think it’s one of the things that’s taught me about life the most. It’s mostly like, for example, you do boxing, you go against your opponent, and it’s in one of these situations where, oh, I got beaten up. And it’s most of the time you realize, well, I’m at fault, right? And this competitive environment, it’s more like a dance between two people more than it is this violent thing. Yes, there’s some risk of injury and something that might happen. Of course, there’s a risk to it, but it teaches you a lot about life and making mistakes and then that it’s you against the other person and it’s mostly your fault if you do a mistake, it’s not the other person’s fault, if that makes sense.
Scott W. Luton (04:48):
It does, Nick. I tell you, I’ve never… Rick, I took karate back in the day. Zen Shodakah was the form I took back when the Karate Kid made all of us go crazy for the martial arts. Remember that back in the day? But I love Nick’s perspective and the bigger picture of what it’s meant in his journey. What’d you hear there, Rick?
Rick McDonald (05:09):
Well, that was not my jam. Sports with a round ball was mostly what I did, but the way I equate it is, I mean, it’s about discipline, both mental and physical. It’s about handling snap decisions. And if you think about what we do in the supply chain, and maybe especially in the procurement space, this really all applies. It seems like a great foundation.
Scott W. Luton (05:31):
I’m with you. I’m with you. And Nick, have you ever had to threaten a chop or a hold or throw in some of those supply chain problem solving conversations, Nick?
Nick Gospodinov (05:42):
No, absolutely not.
Scott W. Luton (05:45):
Only kid. Only kid. All right. So Nick, I bet we could talk about martial arts given your passion for the whole hour here, but we got a lot to get to and the cool things you are doing at Mandel AI and beyond, but I want to level set first. You have an interesting background. Tell us more around that and tell us what got you into solving big supply chain problems. Nick?
Nick Gospodinov (06:07):
It’s interesting question because I do not come from a supply chain background at all. I actually started programming when I was 10 years old and I got into computers, of course, through video games and I was just curious how are these pixels moving on my screen and how can I… I really though that there was probably a person drawing every single scene in the game. I was like, it doesn’t make a lot of sense. And that’s how I kind of got into computers and I never though that I would end up in a space like supply chain, but I think it kind of came naturally to me through just explorations of different ideas. We actually pivoted quite a lot in the early days of our startup, and I found it’s a very creative kind of journey, just getting from an initial very surface level understanding of some kind of industry and then diving deeper and understanding the thing that’s moving trillions and that’s very core to how the world operates is still very, very manual and essentially driven by duct tape.
(07:12):
And so that’s kind of how I ended up working in supply chain. I just though it’s a very, very, very interesting problem to be solving. And the more I explore it, the more we get into it, the more I realize there’s even more broken parts. And to me, that’s way more interesting. So when I hop upon a call with a potential client and they start explaining their problems, I get all excited. For them, it’s like, here’s my pain. It’s like a therapy session. But for me, it’s like I just get even more excited because now there’s a slightly different view of this problem that we’re already solving and it’s amazing.
Scott W. Luton (07:48):
Nick, your passion’s coming out. I think you must exude passion in these conversations as you’re describing. Rick, react to what we just heard there from Nick’s background and what got him into solving these big problems?
Rick McDonald (08:02):
Well, it’s a pretty cool background. I think what it tells me is that if you’re smart, if you’re interested, if you’ve got drive and you want to be a problem solver and figure things out, there are lots of problems in the supply chain, and so he’s on the right spot.
Scott W. Luton (08:16):
Yeah. Well said, Rick, and you would definitely know. Two other things I would add to that is we need this industry that we all know and love needs everyone from every walk of life, whether you’ve been in supply chain or you’re completely new or maybe you’re cross-training fields mid-career. We need all personality types and those from all functions. And then secondly, Rick, gosh, we still, even here in the golden age of supply chain tech, we still have so much manual work and friction that we can eliminate massive opportunities and we need Rick, I would argue, more thinkers like Nick that see those opportunities and better yet are finding ways of eliminating them. Rick, your thoughts?
Rick McDonald (08:59):
Yeah, I agree with that, Scott. And you’ve heard me talk before about the digital fluency of leaders and it’s starting to improve, but I think there are a lot of executives and a lot of senior managers that still need to plus up their capability and their understanding of all levels of artificial intelligence through to Agentic and understand what those technologies, what kind of problems do they solve best? I think we’re going to hear a little bit about how Mandell does that in the procurement space here as we go forward. I
Scott W. Luton (09:26):
Am with you. Digital fluency, yes. That is one of the top three things I hear, Nick, when I spend time with Rick McDonald. Okay, so Nick, you’ve described this intriguing contradiction of sorts, right? As I shared in the opener, supply chains, according to some data, moves more than $26 trillion in goods, yet so much of the communication that makes that happen and holds everything together still happens through email, phone calls, spreadsheets, fax, all sorts of older technology, right? Sometimes I think I heard you say, Nick, duct tape. Sometimes it feels like it’s duct tape, chewing gum and WD-40, right? So I want to ask you, Nick, how do we get here and why haven’t decades of continued innovation and supply chain technology and investment solved this problem we’re talking about here today?
Nick Gospodinov (10:15):
Well, think about it as we’ve been for the past 50 years been trying to structure the transaction. We have EDI, we have portals, we have ERP systems, and we keep stacking more and more data, but we kind of assume that both sides have the same level of, as Rick said, digital fluency. So me and my suppliers need to use the same kind of standard in EDI. We have to have essentially the same kind of ERP systems in order for everything to work actually correctly. And then when it comes to supply chain, the president goes crazy, announces new tariffs, there’s a new war. The world is ever changing, and I think it’s going to continue changing at the same, if not a faster pace. And so how do we keep structuring transactions in the old fashion and continue to be crazy enough to think that this is going to continue to work?
(11:07):
And me and you guys are going to exchange a lot of information here as we speak, and that is unstructured bits of communication. And so we’re going to exchange a lot of information that way, but if we were trying to now structure this podcast after this in some structured set, it would look completely differently like this podcast and the next podcast, otherwise we’re losing semantics about the conversation. And I think this semantics of how an order, for example, evolves from you initially sending it to the supplier and finally getting the goods delivered are probably the most important thing when it comes to that transaction actually executing and the goods being actually delivered in your place. And so these semantics is what we’re working on with Manda, whether it comes through an email, whether it comes through a conversation, but essentially why is there a back order?
(11:58):
Why are we getting overbuilt? Why are we going to have slightly less of that good being delivered in our warehouse, et cetera? All
Scott W. Luton (12:07):
Right, so Rick, he just described, I think, a lot of things we can relate to, especially those in the trenches fighting the good fight every day. But what’d you hear there from Nick in terms of how we got here and why we still haven’t solved some of these opportunities that we’re all through speaking about?
Rick McDonald (12:25):
Well, a couple things. First of all, Nick, you mentioned a couple of macro geopolitical type events. I would argue that in addition to that, there’s sort of the micro level. It’s the working at the speed of the consumer because everybody can use one of these five minutes from now and change a demand forecast if some item goes viral. And that’s where I find this to be the most interesting and the problem to solve is how do we take all these various forms of data, whether they’re coming in by snail mail or fax, and yes, there are suppliers that still do that all the way through to sophisticated APIs and EDIs. How do you take all that and make some sense of it so you can respond at the speed of the consumer?
Scott W. Luton (13:06):
Yep. And what both of you are, or Rick, what you just described there especially, how can we drive this change while not interrupting any orders from going out the doors each and every day, right? It’s a big challenge. Okay, so Nick, I did some homework and I wish I was a fly on the wall when you were having this conversation, but this is what my due diligence told me at least because you’ve been talking with a variety of company leaders out there and organizational leaders, but one in particular I think really sticks out to your journey. You were talking with a Fortune 500 supply chain executive about better planning and forecasting when this leader, I think, turned to you and said, “Nick, the plan doesn’t matter if the latest truth about an order is sitting unread in somebody’s inbox.” So Nick, what did that anecdote or that moment in time illustrate about where the real supply chain information gap exists?
Nick Gospodinov (14:05):
Well, that’s probably the biggest turn in our company’s journey because as I told you earlier, we were pivoting and exploring different ideas in the supply chain space, in the manufacturing space, and one of them was related to forecasting and planning. And there’s quite a lot of tools on the market that promise you a slightly better plan, a slightly better forecast. But once I spoke with the execs, they essentially told us the data in the first place is bad, it’s badly structured, it’s not that clean. And also the biggest problem is it’s not even there. It’s like in somebody’s inbox fits in somebody’s head, and then how do we even structure that? I started asking question because I live in the digital world since I was very young, and I expect everything between two parties to be exchanged over what we call APIs, and that’s not happening in the supply chain.
(14:59):
I thought it was in my mind, it was just the first thing that my mind goes to. It was just not there. And so once I learned about emails, some ideas started turning up in my head and I was like, “Okay, so large language models, AI can now read a lot of these emails, so maybe there’s something we can do here with AI and emails.” We started just talking with more executives, doing some pilots, exploring different ideas, and then we started seeing just how much there is to that email in the first place. And of course, now we’ve expanded to quite a lot more things, but in the beginning, essentially what Mando was doing was just getting emails and getting the latest lead time changes back into your ERP. That was the first thing that we built.
Scott W. Luton (15:42):
There is gold, lots of gold in those email hills, kind of Nick, to your point. Rick, when you hear a digital world native like Nick finding out how supply chain works, and by the way, you’re a great example of that Fortune 500 supply chain leader. When you hear that conversation, that pivotal moment in Mandel AI’s journey, what comes to mind, Rick? Is Nick onto something here?
Rick McDonald (16:10):
I think he is. And the reason he is is because just like human beings where you’ve got five different generations in the workforce, maybe soon to be six, you’ve got multiple generations of technology and all that technology contains some data, it contains some facts, it contains some important information. I did some homework too, and what Mandel does is it helps the humans who are chasing those quotes, who are chasing those missed orders, the late payment, the credits, the off invoice, it’s helping to bring all that together in a much more timely way so you can figure out what to do next at a speed that humans can’t replicate.
Scott W. Luton (16:48):
And just starting and stopping with email, and of course that’s not everything, but Nick, going back to that pivotal moment, just earlier today I was live streaming, had a great show with a couple of chief supply chain officers from big healthcare organizations, and we were talking about this holiday called Empty Inbox Day, and it’s actually a thing, right? Raise your hand if you’ve ever had an empty inbox. Nick, Rick, nobody has, right? Hardly anybody. And it really just goes to this point you’re making. There’s so much actionable information, intelligence and answers just sitting there in our email inboxes waiting for us to find and act on it. So Nick, before I move forward, it’s not that simple, but it really is a big problem out there, huh?
Nick Gospodinov (17:44):
Oh, it’s a massive problem. And I think the biggest thing we underestimate is when I speak with executives, it’s hard to grasp what is the cost of email. You apply AI to email and it’s doing stuff, right? But to me as an executive, email is not really a problem, but what do you fear? It’s like the outcomes that never actually arrive because you don’t have that system in place and because you trust the people. And when you think of email, where are you going to first understand that there is a back order, for example, on some kind of order? Well, it’s going to be email. Where are you going to understand that there’s a shortage? It’s going to be over email. So the most critical things when it comes to the resilience of your supply chain, when it comes to the actual leakages that occur outside of the things that you kind of expect to happen in the day-to-day are going to happen over email or over a phone call or over a meeting.
(18:38):
But essentially, it will be unstructured bits of communication that you never actually capture and it’s in somebody’s head. And then when that tribal knowledge goes out of the door, you’ve actually lost years of development of your company to the so-called email, but it’s essentially just knowledge going out of the door.
Scott W. Luton (18:57):
So Nick, I’m going to revisit the scenario that you were kind of alluding to here in just a second. But Rick, I like how he implied the tribal knowledge piece because when we think about email and we think about the folks that sit in our offices, our warehouses, our manufacturing sites, you name it, especially those with tenure that just know how things operate and it’s right between their ears. And gosh, they go win the lottery and they’re out of the organization or if they retire or if they whatever, that’s such a massive threat for our organization, not just to lose the emails, which is one thing, but to lose all that knowledge and that steady hand in so many ways on our operations. That’s also a massive threat here, Rick, huh?
Rick McDonald (19:37):
It is. And I think this combination of talent and technology is the real key, and it’s a design choice and it’s one that it seems like Nick has made with his organization. I like it a lot. The reality is tribal knowledge, somebody who’s been around for a long time, understands the workflows, the processes. Even still, they will probably not be and likely not be the most capable person to make multiple decisions in a chaotic environment. And if they’re trying to hold things together as a human being and kind of grabbing a bunch of different threads and it’s chaos all at the same time, they will not have a good grasp on how to move things forward and how to solve those problems in an expeditious manner.
Scott W. Luton (20:18):
And by the way, before I move forward, I reference the email conversation. Regine Valley, former chief supply chain officer with Ochsner Health down in New Orleans, she always used an inventory approach for emails, first in, first out. Nick, Rick, I can’t do that. She’s going to have to teach me email 101, but hey, for some, whatever works. All right, so Nick, you were referencing earlier a little bit of a simple scenario here. A supplier sends an email saying half of a critical order is back ordered or maybe the delivery date has slipped a bit. We all probably have been in that situation. I know I have. Someone’s got to go find that email. They got to understand the impact. They got to chase the supplier or have conversations or Zooms or what have you, figure out what to do next. The solution isn’t always all sitting there in front of you wrapped up with a bow.
(21:10):
So Nick, I want to come back to what y’all are doing. What should AI agents be capable of doing with that information and how far should we automate and lean into the automation of what those AI agents are doing? Your thoughts, Nick?
Nick Gospodinov (21:28):
Yeah, so when it comes to our own clients, we always start with human in the loop first approach. I don’t think anybody’s going to trust AI now to go and ultimate a full back order and make a decision and then go back to the supplier and put something in the system and maybe even find an alternative supplier and send an additional order with them. So there’s quite a lot of steps to it, but I think people don’t realize how far AI has actually come in terms of being able to operate these multi-step workflows. So we always start by first educating the people first of what is possible, but always starting from the lowest common denominator and then evolving up to the big workflows. But what can AI do in that situation? So I’ll give you something that we built specifically for one of our clients is they were missing a lot of these backorder notices and they essentially had their own playbook on how to go about handling them depending on the situation, depending on the spend, depending on the materials.
(22:30):
And what they wanted AI to do is in majority of the cases, it sees that order gets confirmed, but some of the things are on back order. And then in that situation, it wants to go see the approved suppliers in the procurement system and then go to these approved suppliers, see if we have a still active quote. So that’s another bit that a lot of companies don’t even know is when you’re running RFQs, for example, you get multiple quotes, but you only store the winning bits. But if you have a system that follows the semantics, captures the full context of every transaction RFQ, et cetera, we actually have more fresh pricing data. So our agent can now go and see if we have still a quote that’s still active, it’s not expired, but we also run an RFQ with the other two approved suppliers and then bring the information back into the buyer’s heads.
(23:21):
And so now the buyer doesn’t need to seek, do we have a problem in the first place, doesn’t need to then make a decision, “Oh, what was the playbook in that kind of situation?” Then actually go execute the operational work to actually do the playbook and then actually make the final decision. With Mandela in that situation, they would just get the final, “Here’s the three polls that I gathered from approved suppliers, and here’s this one other supplier that I found online that could possibly also fulfill this back order. Do you want me to go execute an RFQ with them as well, or do you just want to go with one of the calls that we already have available?” And then it’s essentially a yes/no decision to the buyer, and all the work around that has been assisted by the agent.
Scott W. Luton (24:04):
So Rick, two things I picked up as Nick was sharing there. Number one, man, I think he’s right. AI in terms of multi-step workflows, the capability is here. Our industry’s not taking advantage of it enough. And number two, the checkpoints. I think the checkpoints is probably something that may be different for every team and every organization based on comfort and based on maybe complexity or what have you. It makes me think of my days in metal stamping. We had some very complex accounts and of course metal markets and all kinds of different factors that went into how we price parts and stuff. I wish we had AI back then to bring me not all the detail, a tidal wave of details, especially in a contingency to bring me, “Hey, this is what we did. Do we proceed yes or no? Oh my gosh. Nick, where were you about 14 years ago, my friend?” But Rick, what did you hear there, especially in that scenario that I think a lot of folks listening or watching can relate to?
Rick McDonald (25:02):
Well, it starts and ends with the buyer, which I like a lot. And in the middle is a whole lot of cool technology that accelerates the ability to solve a possibly unknown problem or a hidden problem quickly with a recommendation about what to do about that problem. So this is all the messiness that happens between transactions. And while the ERP is recording the action, Mandel is actually describing what to do next and bringing a recommendation back to that buyer so that they can say grace over it and say, “Yep, that’s something I want to do,” or, “No, I want other alternatives.” That’s a pretty powerful capability. And if you think about what that does, it speeds up the whole error resolution, issue resolution process, and that could have cost implications, margin implications, customer relationship, and reputation implications. Those are all three pretty powerful levers.
Scott W. Luton (25:57):
I’m with you. And just my personal mantra, y’all might not agree. I want more answers, less dashboards. And Nick, I love the very practical solution that y’all are bringing a lot of organizations out there, but here’s a question for you, Nick, and this kind of goes back to those checkpoints we were referencing earlier, and this may vary from organization to organization, at least that’s what I see. The human judgment, right? Where do you see that as being really important to still be involved and also in control? Nick, your thoughts there?
Nick Gospodinov (26:30):
I’ll give you a good example here, or an analogy more so of sorts. I grew up playing real-time strategy games, and when you’re playing real-time strategy games, like Age of Empires or one of these old ones, I don’t know if you –
Scott W. Luton (26:43):
Civilization. You ever play civilization? That’s a good one.
Nick Gospodinov (26:46):
Exactly. And so the human is at the center of the strategy. While you have these agents chopping wood, going to fight in a war or doing any kind of these kind of actions, but still the human is there controlling 200, 300 agents that are actually executing work. Why don’t we have that in supply chain? This is essentially what we’re trying to achieve is can I have one person doing the work of 300 people that would’ve been until now, essentially. So this is the human judgment that’s going to be left. I just think it’s not about replacing, it’s literally about how can we increase the capacity hundredfold, a thousand fold, depending on, of course, how much work there is through AI.
Scott W. Luton (27:32):
So Rick, you were nodding your head there. React to what Nick just shared.
Rick McDonald (27:36):
Well, first of all, I feel like I really missed out on something with a game called civilization. I was busy playing Donkey Kong back in the day. But this is the interesting part of it. The agents are going to work to control points that the client sets. Some of those may have broader latitude that they want to give the agents, some of the tighter latitude. It may depend on where they are in their digital maturity, but the important part is all of that work is happening in the background as the human being goes and does something else that’s important to their day, important to their enterprise, important to their performance. And it is a big capacity release, and I think we’re going to hear a lot more about this as we go forward.
Scott W. Luton (28:14):
Undoubtedly. And before we go forward, Nick, I’d love to just have a follow-up question here because one of the conversations, and Rick, I think you and I have talked about this, I think. I can get some of my wires crossed, but I’ve talked to a lot of folks here lately when it comes to how work is evolving, and then going back, Nick, to the Age of Empires example, where you got armies and armies of folks just doing stuff that all roll up to one strategy or one human control, you name it. But do you see a big need for as work evolves, how we’ve got to reset our organizational structure and strategy rather than just throwing technology on top of the existing strategy so we can take best advantage of what you’re describing, Nick? Your thoughts, sir?
Nick Gospodinov (29:04):
Scott, absolutely. I think it’s, again, as executives have been speaking about 10, 15 years is essentially about, yeah, I want to have actionable insights, but I don’t have clean data. I think now with AI, it’s like, yes, I want actions to be taken across my organizations, but I’ve never read my SOPs or I don’t read them often enough. I challenge every leader to be fair and say if they’ve actually seen these SOPs be executed one-to-one the way that they’re actually described. This comes down to what we were saying earlier about tribal knowledge. We kind of take a fundamentally different approach is where we don’t go about trusting the SOPs as the final deliverable that the agents need to do, but more so, I’ll give you an example. We work with a very, very large outer space distribution company that had hundreds of buyers. And of course the SOP of how to run an RFQ was there.
(30:05):
But when it came to actually executing that RFQ, I remember this one very good example where the buyer was sending out RFQs to this one specific supplier, not as an attachment to the email, but as an inline HTML table inside the email because the supplier on the other side had an alternation of themselves that was taking inline HTML tables and quoting faster. So of course, this is again in relation to what we’re saying earlier is that with the technology so far, I needed to be moving with the pace of my supplier. But with AI, the AI can just remember for that specific supplier that the tribal knowledge is that we send out the RFQ as inline HTML, not as an attachment. And then now I might be sending 50 RFQs and 49 will be an attachment to the email or maybe some will be over a portal and then this one will be an inline HTML.
(30:58):
And then now yes, the SOP and how we run RFQs overall stays the same, but the individual buyer supplier relationship or just company supplier relationship is kind of remembered by the AI. And then now it’s turned into execution.
Scott W. Luton (31:15):
I appreciate you sharing your thoughts there. And Rick, I think we talked about this on the last webinar we did together where you convened one heck of a trio of leaders. Rick, we’ve got to take just like we take a living and breathing and innovative approach to technology, if we’re leaving our organizational structure and SOPs or whatever else goes into that and it’s very static, we’re missing a massive opportunity, Rick. You see that too?
Rick McDonald (31:40):
We are. And I would add, it’s the thing that Nick started with and I’ll add to it, this idea about the quality of data and the reverence that we should all have for our data. It’s arguably the third most important asset a company has behind its people, its brands and products. And I put data right there as number three and you could argue it might even be number two, but the reality is until we do a much better job of controlling what goes into the data funnel, who has access to it, who can edit it, we’re going to continue to get trash out the end of the pipe and that’s going to not serve us well in an era where we won’t have a human in the middle to correct some of that data real time. We’re going to be counting on agents to take the data that’s there, the workflows that are there and then act accordingly against that.
(32:30):
And if the data’s wrong, you’re going to end up with some interesting conclusions.
Nick Gospodinov (32:35):
Rick, I just want to throw a quick challenge here is that I think when it comes to agents, they can work with imperfect data. That’s something that I’ve noticed quite a lot compared to previous technology. The question is what kind of data are you going to give it? And that’s when it comes to the unstructured plus the structured because if I can see imperfect quotation information, let’s give it an example. So I send out an RFQ and I only see the winning bit, but I don’t see the 15 other quotes that I received, then the agent will not be working with perfect information the next time. But if I do give it the email and I kind of build this ontology connecting the pieces between the unstructured and the structured world, then now the agent has full visibility, can go and research. If you actually go into ChatGPT or even call today, you can see what’s the difference between a chat interface and an agent that goes and actually does multi-step workflows or reasoning.
(33:31):
And this is a difference. When we interconnect the two worlds, it can go and then reason it can stitch the pieces between the imperfect information.
Rick McDonald (33:40):
Yeah, I fully agree with that, Nick. And I’m seeing the same thing in the whole demand sensing space where the structured data is one thing, but it’s bringing in the unstructured data that matters to your business, to your brand that really makes a big difference.
Scott W. Luton (33:53):
Good stuff, Nick and Rick. And by the way, Rick, your comment about data and how valuable it is, I won’t name names. And then I know this is going on for a while, but still this one got me. There was an airline company that went out of business and one of the big technology companies struck an incredible deal to acquire all of their data. And I’m talking about pricing, marketing, even employee data, which is going to create some litigation. And so these data markets, it’d be interesting to see where we go to your point, Rick. Okay. Nick, I want to ask you a twist on some of the things you mentioned here because I know one of the things you’ve mentioned is companies can’t simply throw more people at some of these problems we’re talking about, particularly as you’ve already called out the geopolitical volatility and lots of other volatility forces manufacturers and many other companies to change suppliers and also manage increasingly complex networks.
(34:54):
And we also got to face it from a talent and a workforce standpoint. There’s not exactly an overflowing abundance of supply of talent out there. One data point I saw here recently is some estimate in the US manufacturing sector alone. There’s currently about 522,000 manufacturing job openings right this minute, which I think speaks to that point. So all of that, Nick, begs the question, how does AI fundamentally change those economics of supplier coordination that you’ve been touched on a couple times where it gives companies the ability to tackle more complexity in a faster moving world without adding all that headcount, which is really challenging in and of itself? Nick, your thoughts?
Nick Gospodinov (35:42):
Yeah, it’s about, again, just how do I transfer that 60-year-old employee’s tribal knowledge into the head of an agent that can now execute it? So even if I bring a more junior person now in, they can actually go and execute together with the agent something that would’ve taken them three years to learn or maybe even never, depending on what the structure of the company is. And I think it’s again about how can I capture more capacity out of a singular person with that technology? Because if you think of a traditional software service systems, your ERP, whatever, it’s kind of like we were promised this system that does work for us, but actually for the past 30 years since we’ve had this technology, we’ve actually had to hire specialist people that work with this particular ERP, this particular procurement system, this particular whatever kind of software specialist, because it takes so much time to actually work to the point where I actually hire people to work for the software instead of having software work for me.
(36:51):
And I think this is the transition that’s applied with AI. And I think it’s a lot of the companies in the space are also taking a very wrong approach about it. It’s like if you go and see majority of the players in procurement and supply chain AI right now, they’re kind of still applying the same playbooks from the old software world where you kind of think that every single organization works in the same way, which it doesn’t. And what we are striving to do with Mandel is essentially how can we make this system run exactly in the way that you want it to do without me having to throw this consulting big workforce at you and have 10 people come at your location and actually work with you over the next three years to actually configure this big system. We actually aim to achieve ROI within the first 24 hours of onboarding a company through, for example, connecting the emails and then seeing, oh, have you actually noticed that you were getting overbuilt on this one invoice?
(37:50):
Have you noticed that there is a material substitution notice in one of your orders that an engineer has not looked at yet or quality hasn’t looked at yet? And so there’s quite a lot of these kind of things that we can observe at first and maybe we’re going to make some mistakes in the first 24 hours or first few weeks, but essentially through that, the AI starts working with you and capturing the tribal knowledge of how these kind of things work, how your playbooks run, how your SOPs run. And then all of a sudden, the more you use it, the more it becomes better specifically for your organization. And then now you have your own very configured supply chain AI system through Mandel that works exactly like your company wants it to work.
Scott W. Luton (38:33):
All right, I’ll take three of them. Nick, send them my way. Okay. Kidding aside, Rick, what he’s describing, two things that stood out. He shared a lot there that we could have a whole podcast series around, but two things that I really liked is we have leaned on so much of the specialization of how we operate or the technology we use, you name it. And all that won’t go away. I don’t think that’s what Nick is suggesting, but rather than be a constraint of finding that perfect specialist, that constraints can go away a little bit, number one. And number two, prior to that, his comment around taking someone, no matter their tenure, let’s say they’ve been with a company five years, all of a sudden empowering them to operate and to act and to realistically accomplish what someone with maybe 20 years of experience with a company has done.
(39:28):
That is exciting to me because in some of the jobs I’ve held, I wish I had 10 times the amount of tenure in an organization because there was so much to learn and to act on. What’d you hear there, Rick, from Nick?
Rick McDonald (39:43):
My listening, Scott and Nick, and really interesting here, my listening was what we have here is a very flexible tool that can be configured quickly to the norms, the workflows, the identity of the client organization. That’s pretty huge. That’s pretty huge. Some people are all about no customization, don’t want customization, want the standard tool. But I think what I heard Nick say was an approach and a process whereby that becomes very powerful very quickly. If you’re gaining value in the first 24 hours, that is a huge thing. The other thing I heard was, without using the words, this is a lot about the approach the client will use to change manage their way into this capability. And by the way, that is one of the biggest disruptors in digital transformations today is poor change management on behalf of those who are taking on new software and new applications and those that are applying it.
(40:44):
I hear Nick say he’s got a solution for that, which is great music.
Scott W. Luton (40:47):
I agree. That’s why I want at least three of them, Nick. All right, so Nick –
Nick Gospodinov (40:52):
We’ll talk after.
Scott W. Luton (40:54):
What’s that?
Nick Gospodinov (40:54):
We’ll talk after this.
Scott W. Luton (40:55):
All right, good deal. Nick, you’ve kind of laid this vision out through the conversation thus far, but I want to circle back and make sure we’re not missing anything. My impression, my hunch is Nick, you go to bed thinking about this and the impact it can have, you jump up first thing after you turn off the alarm and your brain picks back up that this is what drives you. And this world that you envision of where supply chains coordinate themselves, I love the notion, the idea and how you’re acting on this vision you got. So what else have we not talked about today in terms of what a self-coordinating supply chain really looks like and how Mandel AI is truly bringing that to fruition? What else would you add, Nick?
Nick Gospodinov (41:45):
Scott, I think the biggest thing is coordination, as we mentioned earlier, is not just email. Coordination is how do I take my existing systems, how do I take the actual world and what’s evolving in the world, the news, the emails, the spreadsheets, the knowledge, the internal discussions and everything else into a system that captures both the structured information and the semantics so that then we can turn that into execution. And so if you go to our website, actually we advertise ourselves as an AI coordination layer for your supply chain. Of course, email is part of it. I think in my opinion, the reason we can achieve a very fast ROI is because we connect immediately with email, we can start just searching and then presenting different issues and things like that. But you can see how rapidly you can evolve when you start interconnecting all the different systems and how much coordination can actually become the biggest thing, advancing your company forward.
(42:46):
And supply chain is the foundation of your company. And so see yourself, even if it’s not Mandel, just like if it’s another company, just see yourself five, 10 years into the future and how you can turn coordination into your advantage. So remove all the coordination bottlenecks, getting from engineering to supply chain, getting from production to supply chain, et cetera. And then if we remove these, how much faster my company can operate? How more accurate can I be to my customers? How more accurate can I be with my production planning? How more accurate can I be with my forecasts? It’s actually like coordination is the foundation. So this is the biggest thing that we’re working on.
Scott W. Luton (43:27):
You both mentioned results, outcomes initially inside 24 hours. Nick, that’s really good news. That is really good news. Rick, react to this vision. And it’s more than a vision because they’re already acting on it, I think. They’re already working with companies to bring it to fruition. This is palpable. This can move some mountains. Your thoughts, Rick?
Rick McDonald (43:51):
Yeah, this is a bit of a game changer here because I see Mandel moving procurement from a system of record to a system of action and doing it in a very methodical yet quick fashion whereby you can improve your cost, you can improve your margin, you can improve your reputation, you can improve your customer service. By the way, the next time you go to RFP those suppliers and you become a more predictable customer to them, their contingency, the dollars they put in because your relationship might not be as stable, it could certainly improve. You can see some cost savings there as well because of your reliability. So I just see there across the entire spectrum, I see lots of room for opportunity improvement that Mandel’s delivering.
Scott W. Luton (44:35):
Nick, did you hear that? I love that. I wrote that down. Moving from a system of record to a system of action. Rick, you got to trademark that. I’m going to steal it from you, my friend.
Rick McDonald (44:45):
Take three of those two, Scott.
Scott W. Luton (44:48):
So Nick, I tell you what, we’re going to have to have you back. I’m excited about what you’re doing and I love your passion. And almost, Rick, this is just my take. Nick, you might disagree, but it’s like Nick is so dissatisfied with what he sees out there in terms of how supply chain business happens that he is just insistent that it’s got to change. And I think when I think of some of the most smart and savvy and impactful people that I’ve rubbed up bows with in my journey, it is that constructive dissatisfaction to steal a phrase from another big organization that I think is like rocket fuel. Rick, do you see the same thing or have you seen the same thing in your journey?
Rick McDonald (45:30):
I do. And the procurement space is responsible for, in most companies, the entire supplier community, maybe as much as 80% of a company’s spend overall. And it’s a massive place where the opportunity for productivity improvements and margin improvements exist in spades.
Scott W. Luton (45:50):
All right. So Nick, for those out there that are listening or watching and your story and your passion and what you are doing at Mandel AI, they want to learn more, how can they connect with you and your team, my friend?
Nick Gospodinov (46:06):
Scott, everyone can find me, either sign up on our website or reach out to me at nick@mandel.ai. It’s very simple. It’s my first name @mandel.ai, and I’m frequently on LinkedIn. You can find me on LinkedIn. We’ll be very happy to talk with all of your viewers.
Scott W. Luton (46:25):
Outstanding. Well, big thanks, Nick. Appreciate what you’re doing. Look forward to having you back soon. Rick, before we sign off here, we’ve covered a lot of ground with Nick and what they’re doing, the change they’re driving. But Rick, if you had to narrow it down to one of your favorite key takeaways from what Nick has shared with us here today, and in a broader sense, the opportunity that’s out there that you’ve spoken to a couple times already, what is this one takeaway that our SEN Global fan have tuned into this episode can’t forget if they forget everything else? Yeah,
Rick McDonald (47:00):
I’m going to merge two things into one. First of all, the human in the loop to make the next thing happen. Taking all of this data structured, unstructured and making the next thing happen with the human right in the middle of it. That’s a very powerful concept.
Scott W. Luton (47:15):
I’m with you. And folks, we heard from Rick McDonald who went to the Gemba for this conversation here today. He’s in the warehouse where the action and the value is created. Rick, thank you my friend, Rick McDonald. Always a pleasure to have these conversations with you. Thanks for being here, Rick.
Rick McDonald (47:31):
My pleasure, Scott. Thanks for having me.
Scott W. Luton (47:33):
And Nick Gosporinov. I have really enjoyed your perspective here today, co-founder, CEO of Mandel AI. I love folks that are really, really passionate about changing how things are done, especially we got this beautiful human workforce out there that drives our industry and they work so hard every single day. And unfortunately, there’s a big portion of what they do that we’ve got to upgrade so it’s easier for them because they want to be successful, they want to challenge. And sometimes we can all think like it’s 1988 and we got to change that. And Nick, you’re on a mission to change that. Thanks for being here, Nick. I look forward to seeing what you and Mandel AI are up to next.
Nick Gospodinov (48:17):
Scott, Rick, thank you so much and it was a pleasure to be here.
Scott W. Luton (48:21):
Same, same, same. All right, so Rick and Nick, what a great conversation. To our SC and global fam, I hope I’ve kept those two names in their proper swim lanes because my brain, I was crossing wires regularly, but I thought I really enjoyed Nick and Rick’s perspective here today. It’s very actionable. It’s very actionable and I hope that y’all know your homework, but that’s why we strive to really create actionable conversations. You got to take one thing from all the good stuff you heard here from Nick or from Rick, take one thing, do something with it. Deed’s not words. That’s how we’re going to continue transforming global supply chain, global business and leave no one behind. And with that said, Scott W. Luton here from the Supply Chain Now team challenging all of y’all to do good, give forward, be the change that’s needed, and we’ll see you next time right back here on Supply Chain Now.
(49:09):
Thanks everybody.
Voice Over (49:12):
Join the Supply Chain Now community for more supply chain perspectives, news, and innovation. Check out supplychainnow.com, subscribe to Supply Chain Now on YouTube, and follow and listen to Supply Chain Now wherever you get your podcasts.
