Inside the Tech Stack Behind the Modern Store with Yevgeni Tsirulnik & Naresh Keswani of Toshiba

(Intro Narrator at 00:00:00) Today, we're talking to Yev and Naresh from Toshiba Global Commerce Solutions about all of the latest technology in the modern store and how they're making the business case for it. You're listening to Joel Beasley, Modern CTO.

(Joel Beasley at 00:00:19) How we doing, guys?

(Naresh Keswani at 00:00:20) Hey, Joel.

(Yev (Yevgeny) at 00:00:21) Hey, Joel.

(Joel Beasley at 00:00:22) So right at the top, I just want to start with some introductions, who you are, what you do at the company. Yev, could you go first?

(Yev (Yevgeny) at 00:00:31) Yeah. My name is Yev, Yevgeny, but everyone calls me Yev. I lead our software line of business for Toshiba Global Commerce Solutions, or TGCS. I basically cover all of the journey from ideation to the product to innovation, sales, and delivery of the solutions. So I've been in retail for over two decades, closer probably to twenty-six, twenty-seven years. Enjoying it, loving it. Thanks for inviting me here.

(Naresh Keswani at 00:01:07) And then Naresh? Naresh Keswani. I lead the software portfolio in the software business at Toshiba Global Commerce Solutions. My responsibility spans from product strategy, portfolio, ideating with the customers, and thinking about the capabilities that help retail modernization and scale globally. A lot of my focus today is really around Elera. That's our next generation platform. And increasingly, how AI, edge intelligence, and agentic AI fundamentally change how retail software is built, deployed, and consumed globally. So in a lot of ways, Yev, who's always thinking how retailers need to operate in the future and in fact loves to break things, I need to think about how do we productize it, commercialize it, and then bring some guardrails around it. So it's been fun doing that for a little over nine years here at Toshiba.

(Joel Beasley at 00:02:18) Oh, and how long have you and Yev been collaborating together?

(Naresh Keswani at 00:02:21) Four of those nine years.

(Joel Beasley at 00:02:22) Oh, so you guys know each other. It's been a while. Yeah?

(Naresh Keswani at 00:02:26) Yeah. Oh, very cool.

(Yev (Yevgeny) at 00:02:27) Yeah. So in other words, from what Naresh said, I am the one that's trying to think about the future and all the crazy ideas. Naresh is grounding and balancing, and we've worked together for all this time very successfully.

(Joel Beasley at 00:02:44) Yeah. I think last year when I talked to Toshiba, they were talking about how previously it was heavy investment into digital, but now there's a large investment into the physical stores. Is that still happening?

(Naresh Keswani at 00:02:57) That's fundamental, Joel. In fact, if anything that we've seen, and like you just brought it up, right, I think the industry has spent the past decade digitizing the consumer journey, and what we are seeing is a tremendous shift in actually making the physical retail get digitized now. And right, that investment that was all around ecommerce, all around digital journeys, all buy online, pick up in store, and returns modernization, was great. At least in a lot of spaces that we operate, 90 plus percent of either shopping or the influence towards that shopping is happening in the physical store. And retailers are realizing that the modernization of those platforms of the physical retail is paramount. And even if the systems aren't broken as is, they're preventing them from making meaningful business decisions and driving higher revenue unless that investment is made. And that's what, to Yev's point, we've been doing successfully over the last few years is inspiring retailers with where that modernization in physical space needs to happen and how do you connect meaningfully and profitably the offline and online world?

(Joel Beasley at 00:04:23) Are most people, like most of the retailers, are they hungry for this? Are they lagging behind? Like, what are you seeing a lot of out there?

(Yev (Yevgeny) at 00:04:31) The marathon for the digitalization of basically know everything about the customer, know everything about the shopper was all around the bet that ecommerce will take way more share in sales and in retail in general. So it was perfected. They know everything about us in ecommerce retailers. Yet the store is where everything is happening. So to your point, those retailers are realizing that they can't take 10% of their sales and apply the same rules overall in the physical. So they have to go and know, basically they have to know the physical, not less than they know today the digital world, or in other words, connect those things. So to me, retailers are basically going through that evolution of connecting the two worlds right now, getting these two businesses really working closely together.

(Joel Beasley at 00:05:41) That's interesting because if I were to have a conversation with either of you about heat map tracking on my website, click conversion, how long they stared at a certain section, whether they clicked the thing a bunch of times and it was broken, all these triggers, you'd be like, yeah, we've been doing that for a decade. That's nothing new. That's been around, and that's well understood. But are you saying that we're gonna see parallels to that in the physical environment?

(Yev (Yevgeny) at 00:06:07) Absolutely. And that's the key because, again, today, all of the paradigms that they see and all of the data that is coming out of the digital is kind of this prism being applied to the physical store, and it's not always working. So in order for that to really happen flawlessly, they're gonna have to go and do exactly the same what you mentioned for the store. They need to know everything about the shopper, but without knowing who is that shopper. Right? So that's where all of their PII or personal identification data is being protected. Right? So all of that balance between those things have to happen right now in the stores. So yeah.

(Naresh Keswani at 00:06:53) Yeah. It will be—

(Yev (Yevgeny) at 00:06:55) The technology today in the physical store is allowing you to basically do exactly the same thing as you do today on the website.

(Naresh Keswani at 00:07:06) And, Joel, to your point, right, everything that you kind of pointed out, dwell time, click-throughs, previous action, in fact, predicting the next action. The modern thinking is this digital twin in the physical environment. Right? If Joel's shopping with me on my ecom website, third party, uses the app, I'm able to get to them through coupons, but I walk inside the store and the retailer's a little bit blinded by, hey, what prompted you to come into the store and what are you doing inside my store? But with modern platforms invested in and then modular solutions like computer vision, like sensors, with more camera vision coming inside the store, and then connecting the data from third parties. If I can get to a point that I recognize, not exactly who you are, but what your persona is, what did you do before walking in, and if you're coming through the aisle, am I able to now make your shopping behavior and experience get tailored and personalized to you, and then your checkout experience is personalized to you? That's all happening because I'm getting similar signals and data from the store I am used to getting from the digital world. And to Yev's point, if I can bring all that data together and now prescribe your journey, in a lot of ways now as a retailer, I'm in better control in the physical store than what I was in the past.

(Joel Beasley at 00:08:44) That's interesting. Yeah. When you walk in the grocery store, they're blind. You know, you don't bring your cookies with you to the grocery store.

(Naresh Keswani at 00:08:53) You don't. But if I can get you in the aisle and my customer display is able to comprehend because you're connected to my Wi-Fi, I can get pretty dangerous in terms of ensuring that I understand which coupon to send you or navigate you to the aisle, which I think, you know, if you are interested in a wine, can I incentivize you to buy some cheese at 50 cents off? That's certainly I can do inside the store.

(Joel Beasley at 00:09:25) Oh, so people in the retail stores are using the retail Wi-Fis, and then they're connecting the digital device IDs back to their profiles?

(Yev (Yevgeny) at 00:09:35) It's one of the choices. There are a lot of technologies even without any Wi-Fi connection to know where the devices are in the store that potentially can connect the signals. Right? So yeah, it's a solution or that technology has been there for some time. So yeah.

(Joel Beasley at 00:09:58) I talked to someone a few years ago. I'm gonna be wrong here, but it was some store we all know. And I was talking to one of their technology leaders, and they mentioned something along the lines of, we can dictate product placement within the store by looking at where the device ID is, like, stop. Like, we can see when they're in the store, and we can map them. And we can say, like, oh, they go in front of here and look at this, and then they go to checkout or they don't. And so that was maybe four or five years ago they've been doing that. Right?

(Yev (Yevgeny) at 00:10:28) Yes. And even think about the video, like the normal CCTV, even the analog data, that all can be connected to exactly the same use case or the same purpose. So, absolutely, that is one of the easiest ones, and that's what determined the dwell time. And I think also about all of the CPGs, all of their vendors that provide the products to the store, how much they want to know what is the most effective shelf in the store, what aisle I need to be in to be successful in. Right? It's a lot of data points that you can connect. And by the way, that data game has existed for years, for decades. What makes today's world much more interesting in this is all of that GenTech that can very quickly connect those dots together and come up with the insights that usually you would wait for weeks after all of this big data is being consolidated and analyzed to give you those insights. The customer's not in the store anymore.

(Joel Beasley at 00:11:39) Yeah. So to kind of wrap this section up for you guys as a provider of this technology, you're seeing that you're not having to really convince the retailers of, like, replacing their systems. They are coming to you. They're like, hey, they're seeing the results, and they're coming to you largely for these types of technologies.

(Yev (Yevgeny) at 00:12:02) And, again, we have pretty significant share of the market. So we always have these innovation discussions. But to your point, the overall, the discussion always starts with what is the problem we need to solve. So, yeah, sure, we have a lot of foundational technologies in place, so it's much easier for us to evolve it in various use cases. But it all starts from what is the use case? What is the ROI you're trying to achieve? What is the outcome you're trying to achieve? And then we walk back into what technology needs to be used, and does it make sense to use that technology for this?

(Naresh Keswani at 00:12:42) Yeah. Business case is needed no matter what. Right? Even to Yev's point, you need to work together to identify what is the problem we are solving. Is it a customer experience problem? Is it an operational efficiency problem? Is it a loss of revenue problem? And together to build a business case. Are we going to do something modular that helps retailers realize the revenue and or loss quickly? Or on the other end, how do we build a path for a complete modernization, if you will, while realizing revenues on the path. But to directly answer your question, it's a process to work those business cases, but things are moving so fast. Right? What used to be, let's go build something and then show a proof of concept in six to nine months and then get to a pilot. Those things are now happening in weeks, and that's all because of, right, the investments that we've made and others have made in AI and how we can experiment and move forward faster.

(Joel Beasley at 00:13:51) Yeah. And just the technology that's available to all of us. I mean, three years ago, AI was not what it is today as far as just even the normal communication stuff between, you know, business meetings, booking stuff, and just interacting with the brand, even outside of your core technologies. We have so many tools now to speed things up. You have—I read you had an article about moving from modernization to execution. My chief complaint with that was, what's the difference?

(Yev (Yevgeny) at 00:14:24) So I'm gonna tell you my philosophy around this. Modernization, usually people take it, okay, we're gonna try something out, and they can stay in this trial forever.

(Joel Beasley at 00:14:37) Oh, okay.

(Yev (Yevgeny) at 00:14:37) It's just not going anywhere. It's not scaling. Well, it's nice. Yeah. It can be my flagship store. It can be something I can show off as far as the technology. But execution, when I say execution, that means you scale it. Now we're working in retail. Our smallest clients are twenty, thirty stores, which is pretty large for normal people. Right? We have clients with thousands and thousands of stores. So how do you take that modernization and really become, or take it into the scalable, repeatable, economically viable type of solution. How do you get ROI out of it is where I see the difference between execution and modernization landing.

(Joel Beasley at 00:15:24) Are there some people that think, like, hey, I have the loyalty system. I'm good, and that's all they're doing, or are they all pretty modern?

(Naresh Keswani at 00:15:35) When it comes to loyalty in particular, and we can go around modernization in general, right? Everybody's got a system in place, and there's always improvements that can be made. But the bigger question, Joel, is really around the business case for that modernization, and loyalty is just one aspect of it. Right? Certainly, very important aspect of creating that brand value, that connection with the customer, help them come back for increasing of their basket size. But to think about, hey, I've invested in loyalty system, and it's good enough. And if it ain't broke, don't fix it. That's not good enough because if you are not competing and constantly making improvements, somebody else is. Right? And in today's environment, I'm sure you do. Right? You look for an experience at a retailer. You have one bad experience. You're forgiving. You have a second bad experience. That loyalty is out the window very quickly. So you have to constantly demonstrate the value and then keep the faith with the shopper. And it just doesn't belong in loyalty. It comes down to your point, how am I constantly not only modernizing, but then also figuring out execution? Because an existing platform may be just fine and it's not broken, and if you have to go replace it with the next platform, it may be not as cost-effective either. But what makes it more cost-effective is to add new capabilities or do new use cases around it. That's where the new business value comes in. So to your point, if loyalty system X lets me do X today, there are other things that I could be doing that I should invest in, and that might mean that the cost of the next platform might be higher. But if it lets me create new value for the customer, lets me do new use cases, and drive a higher basket, then I would look at replacing my legacy system with the new systems. And that applies not only on loyalty, but mobility, in-store solutions, be it point of sale, self-checkout, you know, outside the store, off-campus sales. Like, there's so many new touchpoints and ways of helping retailers say it, even returns modernization. Right? Those are all investments that we are seeing retailers make in order to improve that shopper experience.

(Yev (Yevgeny) at 00:18:17) You know, the very simple way to answer that question is, you know, when you buy a car and you don't drive the car, why did you buy the car? If you have a good loyalty system that you invested in and your clients are attached to this, why have you done it? You've done it in order to get more value out of this. And, obviously, it's the value both ways. It's a win-win.

(Yev (Yevgeny) at 00:18:44) Right? So if you don't do anything with the fact that you have loyal clients through your loyalty system, what was the goal of doing that? So to me, it's a very simple way of looking at this. Loyalty is there to bring the value to the customer, but obviously to the business as well.

(Joel Beasley at 00:19:05) Yeah. You know, I have never been one for the loyalty programs. I had gone to this one specific grocery store for basically my entire life, and then we moved to a new town. And we tried this other grocery store named Kroger, and they had a loyalty program, and I just rolled my eyes at it. And my wife looked into it, and oh my, it's probably the most effective loyalty program that I've ever seen in my life.

(Joel Beasley at 00:19:32) And now we go there because of it. And, you know, I'm not, again, big on the loyalty programs except for when it comes to this one grocery store because it is so valuable. And then the other thing that I like about that store—and I'm just sharing this as a consumer to see how I relate to other consumers, because what's important to me may not be important to people at large in society—but the quality and price, the biggest things to me, followed by the speed. So if a store doesn't have self checkout or they don't have enough self checkout lanes and they're hanging on to the past, as what I call it, I just don't usually go there because I want speed. I want speed in and out. You know?

(Yev (Yevgeny) at 00:20:16) Yep. So first, yes, it's the price, it's the speed, it's all of that. But imagine you connect the technology to this, and you make the experience completely different for that loyal client. Right? All the way to, you know, walk in and walk out of the store with no cashier, specifically because you are a loyal customer. Right? Basically, because I know who you are, because I trust you, I can give you a completely different journey in the store versus a person that I don't.

(Naresh Keswani at 00:20:48) Yeah. I mean, that's where a lot of investments go in. Right? Even connecting what we talked earlier, this data is now so valuable. Right? Sticking with the loyalty theme, Joel, if I know your purchase patterns, I've interacted with you at self checkout, I have—now there are overhead cameras in all of those stores that you shop at, which we actually proudly invested in AI several years ago. So I have enough data that I have not seen you steal from me, and I have evidence. Right?

(Yev (Yevgeny) at 00:21:27) Because you've—

(Naresh Keswani at 00:21:27) You've purchased. And anytime you are making a mistake, I know I need to forgive you because you are a loyal shopper. And I can—interesting point—give you a completely unique experience based on your past behavior, based on what you're doing with me. Versus somebody else, I will send somebody in to intervene because I need to ensure that there wasn't some behavior that I wouldn't want to happen in my self checkout. By doing that, right now I created a better experience for you, and I reduced the burden on my shopper assistant to come in and interrupt you as well. So I used my employees in a much better format as well. So you can see how that data inside the store is becoming so powerful, almost bringing a digital experience.

(Joel Beasley at 00:22:21) Yeah. I saw the overhead cams for the first time about maybe two weeks ago. My wife asked me to pick up a prescription, and so I went into the pharmacy, picked up the prescription, and then I got some food and then did the self checkout. But I put the prescription bag right in, bypass the scanner, and it replayed it to me on the screen. And then when the attendant came over, I was like, oh, it's so cool. You can catch the—I didn't realize the technology was there where it can track your hand and watch you evade the scanner and put it into the bag and then replay it back to the cashier assistant when they come over. That was neat.

(Yev (Yevgeny) at 00:22:59) And it keeps getting better and better, this type of technology. So yeah.

(Naresh Keswani at 00:23:02) Yeah. It's front-end technology. What's going on in the front end? Did you leave something under, you know, under your cart? And then you can continue to move in the other areas of the store with a similar technology.

(Joel Beasley at 00:23:16) Or your kids. I, one time, was walking out to the car, and my kid had a snack or something they had grabbed off the thing. I was like, what? And so we had to go back in and pay for it. But other than that, we would have just accidentally taken it. You know? Yeah. So I am in North America. I'm in Nashville, Tennessee area. If I were to go experience any retailer—and they're the most advanced one just so I could walk in there and be like, I talked to you, I'm in the rush—and this is the most advanced retailer near me, what would that be?

(Naresh Keswani at 00:23:50) Everybody's got, you know, their own differentiation going. Right? Some that you may experience, some you may not, but they're making investments in the back end to ensure that you get a better experience. I mean, to name a few, I think experiencing the modern carts—that's a pretty good experience if you had. And I don't know if this is where, you know, at least my opinion, experiencing like a Dash Cart is a great experience to see where that thinking, where the future is going, where you pick up an item from the shelf, you scan it right on the cart, you place it in the cart, and at the end, you can pay at the cart and then walk out. Right? You're not even standing in a self checkout lane because you're essentially walking inside the store with the self checkout. And in order to make something like that happen, you're essentially bringing the point of sale to the cart itself. So that's a phenomenal experience.

(Naresh Keswani at 00:24:36) I would also say any of the experiences at a Wegmans store, if you can go visit. That is a retailer that, you know, certainly we proudly support among many other retailers. But going and experiencing their aisles, their inventory that's there, and their accuracy level of inventory. What they've also invested very heavily is a differentiation between a shopper. And in today's economy, you're seeing a lot of, like, Instacart shoppers come inside the store. And that's what I'm saying, Joel, you will not experience, because in any other store, you might see, hey, I'm shopping, but then somebody else is shopping inside the store on a third party's behalf. What they've created is a completely unique experience that if you are a third-party shopper or Instacart shopper, you're adding items from the shelf in your cart, and essentially you're pushing a button and you're sending that order to the cloud, and you have a separate lane at the back of the store that you show your mobile device, they can audit it, and you're out the store. And what that does is it keeps the self checkout and POS lanes open for you. It also keeps an inventory on track, and it avoids unnecessary queuing at these lanes. And that's why, to your point, speed, price, and quality, they're all maintained as you get out.

(Naresh Keswani at 00:26:33) And, oh, by the way, they have this great technology that you should also check out at Wegmans. It's called produce recognition. So if you go place an avocado, you press a button, it's going to recognize it's an avocado. It's going to give you an option of organic or not, and you press the button, it's going to add to your cart. Moreover, if you add bananas that have organic label around it, it will recognize it's an organic banana, and it'll let you add to it. So there's no longer flipping through pages or looking through screens and adding an item. It improves—I mean, we have data. Right? It improves three to five seconds in every transaction. So once again, they've done some phenomenal work in improving that experience for the customer.

(Yev (Yevgeny) at 00:27:13) I can go actually farther on. Naresh is basically talking about the things that some of the stores you can see even today. Further on, I'll talk about two different scenarios or use cases. One is very physical. One is kind of a good combination. You know, there are all these lane lines in the store, the queueing things that usually is done. There's a person comes to you. Imagine a combination of technologies of the self checkout and what we're all used to at home, like Roomba or some other things, you know, those vacuums that know to map your store—your home today. A combination of these technologies where the checkout really drives to you. It's a robot that drives to you, and it has a very little surface where you put your three, four, five small basket items, and it's automatically processed here, and you just don't need to stand in any queues. That's the physical one. And that same technology can take you through various concierge type of tasks in the store. Hey, where is that product? What do I do? What is the price of it? You know, you're looking for the price. You're looking for where to check the item price. That thing comes to you and you just check with this thing. Right? That's the physical one.

(Yev (Yevgeny) at 00:28:24) The more intriguing to me, we're all using ChatGPT these days, and that is already in our lives, probably for over 50% of people today, one way or another using that. So ChatGPT knows everything about you. It knows what you like, what you do, connecting all sorts of things at your home today. You do all of these connectors, and it's becoming a much more powerful sort of a system. Imagine that this agent, basically your home agent, does the shopping for you. So instead of you trying to punch things into this Instacart or whatever the application is there, there is an agent that talks to the agent out there, and your own agent knows how to substitute your items, knows what you like, what you don't like, makes the transaction, finalizes the transaction, maybe asks for your approval at the end of the day, but that becomes your basically your digital robot. That will come to physical world as well at some point. It's just later on.

(Joel Beasley at 00:29:30) Whoa. So I'll just be able to talk to it. It'll learn my preferences over time for replacements and all of that. And then I could just say, hey, order some food for the week, and it'll be like, are you out of peanut butter? And I'll just check. Or it'll have a camera in my pantry. It'll know if I'm out of peanut butter.

(Yev (Yevgeny) at 00:29:48) What type of dinner you want today? Yeah. Or how many guests you're going to have in the next week, and then arrange, surprise me with a nice dinner set.

(Joel Beasley at 00:30:00) Yeah. Now when my wife asks me for dinner, I just ask Gemini. I'm like, what's for dinner? And I just read her whatever. She hates that when I do that. But she asks me what I want for dinner, and I'm like, I don't know. I just ask Gemini. What's a good idea?

(Yev (Yevgeny) at 00:30:14) You know? And that technology is there today. Yeah. What is the problem we're trying to solve, which is solvable, is opening up our platform, which is serving the store today, opening up for all of those types of interfaces so that agent can really talk to the agent without human interruption or minimum human interruption.

(Joel Beasley at 00:30:37) That's so cool. Do you guys have a place for the retailers who are listening to this? Do you have a place where they can come see this technology? Like, do you have a showroom or anything like that?

(Naresh Keswani at 00:30:47) We do. We do. Yeah. I mean, Toshiba Global Commerce Solutions, our global headquarters is in Raleigh, Durham, and we've got a great innovation studio that we host. Retailers come in. We also do virtual demos from our innovation studio for around the globe, actually.

(Joel Beasley at 00:31:08) That's pretty cool. Are you based there, or do you just fly in and meet the customers when they go there?

(Naresh Keswani at 00:31:12) Yeah. No, we're—I mean, we've got centers around the globe. These innovation studios certainly are in Raleigh. I mean, we are—we're there all the time. But there's other places in the world that we have similar centers where we host retailers. Example, Europe. We've got one in—a couple in Asia-Pacific as well.

(Yev (Yevgeny) at 00:31:35) Our teams are all over the place. So Naresh has teams in Raleigh, right, in Frisco where I am. I mean, we've got many technology or demo places pretty much in that office.

(Joel Beasley at 00:31:50) Oh, nice.

(Yev (Yevgeny) at 00:31:50) And there are over 30 of those. So—

(Joel Beasley at 00:31:52) Right. Agentic AI, is that inside of any of your products currently or any retail operations today?

(Naresh Keswani at 00:32:00) I mean, that's all we're thinking, talking about, and thinking what happens in the future. Right? I think it's critical that we first define what does it mean. Right? As I mentioned earlier, we've been doing AI, per se, for five-plus years. Right? And so it's building out modern platforms, you know, getting data. Everybody has models. I think it comes down to, based on that context, what is that agentic AI able to do for you? Right? Even using the example that I gave you earlier. You know, recognizing a product or a shopper's behavior, that's AI. What inference do you drive from it, and what agent can change a workflow? That's agentic AI. Right? Take that example. If I have to understand what's going on at the shopping behavior at the self checkout, what was the past behavior, is there a weather pattern, is there an event that's going on that I need to change the workflow? That's agentic. Right? So those are the type of areas that we are very focused on.

(Naresh Keswani at 00:33:19) But most importantly, our focus is to provide retailers with the core foundation. Right? In fact, we always think about this as three layers. Right? You need a platform. You need a modern platform, be it commerce, be it, you know, data platform, inventory platform. You need to have that. Right? Then it comes, what other data sources are coming your way. So how do you manage that data? What governance do you put in front of that data? And then on top of that is the experience layer. That experience layer could be a touchpoint. It could be an agent. And our understanding and our business proposition is we'll work on all three layers. But the most important thing is that top layer, Joel. I think our retailers want to go build those agents. They want to be able to go do that themselves or bring third party. So our focus really is around how do we ensure the underneath two layers are ready for that agentic AI commerce.

(Joel Beasley at 00:34:25) What do you see as the biggest obstacle from going from experience, like, an experiment to a fully governed product?

(Yev (Yevgeny) at 00:34:36) Specifically to agentic, or you're talking about generating a solution?

(Joel Beasley at 00:34:41) I'm curious about it. Does it matter? Like, are there differences between the failures to get agentic from experiment to a governed product versus just general?

(Yev (Yevgeny) at 00:34:51) That's what we're trying to figure out right now. Product, it's very—if it's a normal use case with the, let's call it a pre-agentic type of solutions, it was very deterministic. It works or it doesn't work. Right? You know what needs to work, you know, and if it doesn't work, you know exactly the flow of what doesn't work. In an agentic world, similar as every one of us experiencing today, AI is hallucinating. And that's happening, you know, pretty frequently for all of us and working around this in your normal life. You just use your own judgment, right, when you're getting an answer from ChatGPT or Gemini or anyone else. And you say, oh, that looks right, or validate this for me or this for me. Agentic in the fully automated world, you have to reduce down to zero any sort of hallucination. And you do that through the governing bodies or agent, the test agent, and things like this. So to your point or to your question, taking it to the fully scalable solution that you can trust is the next challenge that everyone needs to resolve. That's not a Toshiba-only problem. This is a problem of everyone, including the giants that are creating the foundational technology itself.

(Naresh Keswani at 00:36:18) If I could add, I would say if you talk about a general experiment that we've seen over the years in scaling, I think that was a lot clearer. Right? You come up with the hurdles and say, I'm gonna go do this POC. I'm gonna try to meet either this cost number or this revenue number or these KPIs that you define. And then you get to that point, and then you have to think about what does scaling involve.

(Naresh Keswani at 00:36:45) Right? Do my costs scale the same way? That's where going through a faster experimentation, be it pass or fail, that was everybody's drive. And I think we've come a long way across the industry, especially with retailers, that we can move through those experiments fast. In this new AI and agentic AI world, the new wrinkle in all of that is defining those key KPIs and repeatedly showing that we can get there.

(Naresh Keswani at 00:37:18) Right? Because it's just no longer a pass or fail game. It's we know it's not 100%. Right? So all of us to get educated and be able to build this new tolerance towards scaling, I think that's where we are all headed towards.

(Naresh Keswani at 00:37:35) Right? You see a lot of buzz in the industry. People come with new experiments and they put it in production while it's not behaving 100% of the time. The reality is it won't. Right?

(Naresh Keswani at 00:37:47) Even when we ask those questions, like Yev is saying, we don't get 100% answers from AI. Even simple addition problems are sometimes messed up. So that the added challenge that we are all trying to solve is what is that new level of tolerance that says now we can scale.

(Joel Beasley at 00:38:06) Now is this, tell me about the Valera platform. You mentioned it earlier in the conversation. What exactly is that?

(Naresh Keswani at 00:38:13) Yeah. No. I think we've been living Valera for the past six plus years. If I could start out with the problem statement in the marketplace. Right?

(Naresh Keswani at 00:38:24) When we started to even thinking about Valera was, once again, what's the problem that we are trying to solve? And the problems, I would say, right before COVID era, and it continued to be that, if not became even more important, is how quickly can retailers move and implement new innovation? Be it a new business process, new experience. And then how do you reduce cost of aging technology and technology stack, which became a lot more proprietary and very hard to manage. That's when we decided to invest in Valera. So simply speaking, Valera is an open commerce platform built for modernizing both in-store and digital technologies.

(Naresh Keswani at 00:39:22) The solutions that we offer off of Valera platform range anywhere from point of sale, self-checkout, mobility, loss prevention solutions, payments, loyalty, promotions, all things that you need to operate in a transaction environment. But what's not important is the end touchpoints. It's really what's important is how open the platform is for retailers and partners to be able to collaborate and build out new experiences. And that's where our investment has been. And then now bringing more AI and agentic AI capabilities into that open platform, so both partners and retailers can go build out their own experiences and, at the end of the day, deliver value to shoppers.

(Yev (Yevgeny) at 00:40:12) I will just add, Valera very nicely addressing the challenge that I don't think we could have predicted five years ago or seven years ago when we built Valera, which is in today's technology world, everyone pretty much can build something, and the cost of development is much cheaper than ever before. Valera brings both the foundational stability, the scalable part of the solution, as well as the ability for someone to extend it the way they want it, and we're very friendly for that. That is what positioning the Valera solution as sort of a hybrid answer to the question. I can develop myself, but can I assure scalability, performance, and SLA? That hybrid between the ability to extend and very stable foundation is what sells Valera today.

(Joel Beasley at 00:41:12) So if I'm a retailer, I've got engineering people on staff for various reasons, and then I'm using your software, they can, if we have an idea, they can just go write it up and test it?

(Yev (Yevgeny) at 00:41:24) That's exactly right.

(Joel Beasley at 00:41:25) That's pretty awesome. And you guys won an award for it, didn't you?

(Naresh Keswani at 00:41:29) Yeah. Yeah. No. I think that was the core foundation. Right?

(Naresh Keswani at 00:41:33) As I said, one of the key requirements as we jumped into this was the openness. Right? It was being able to be as agnostic as possible, be it infrastructure, the hardware we need to run on, the cloud platforms we need to run on, giving retailers access back. Like you mentioned, if you need to add an experience, you should do that. And for us, we've been recognized across the board as being that open platform or multicommerce platform where, you know, building out a marketplace where we've got retailers who are adding their value.

(Naresh Keswani at 00:42:09) We've got dozens of partners who get access to our platform, to our APIs, and they're able to transact. And then for the retailer, they don't have to go and certify and do commercial agreements and ensure compliance and ongoing maintenance. That all happens through our platform. That's, again, a validation of the openness of the platform that we conceptualized, and it's now in production across the globe.

(Joel Beasley at 00:42:39) If people wanna learn more about these systems, where should they go?

(Naresh Keswani at 00:42:42) They should go to our Toshiba Global Commerce Solutions website. And certainly, both Yev and I, our teams are available on LinkedIn and various social platforms. But TGCS or Toshiba Global Commerce Solutions, look us up.

(Joel Beasley at 00:42:59) Awesome. Josh will post links in the show notes. And then just as we start to wrap up, I always like to end with a leadership question. I'll start with you, Naresh. What's one of the defining moments in your career where you looked back and you're like, okay, that, I changed the way I thought about things. I've really grown as a person. Like, what experience was that?

(Naresh Keswani at 00:43:24) Yeah. No. I've spent a lot of my time in my career in product management. And you would know there's no such stream as product management when you go to college. So you almost have to, you know, you're either coming from engineering or business and you're combining the two and starting to wear that hat.

(Naresh Keswani at 00:43:46) So I've spent a lot of time talking to people, as well as personally going through the various levels of being a great product manager. And I would say the moment that I realized taking on that role that while you're always taught, hey, stay in your lane and perfect what you are doing, I think what was defining for me is for me to really understand the problem that I'm solving as opposed to, no, this is my role, and I need to stay in my role, and I'll depend on somebody else to finish that off so that the problem is solved. I think that's the advice I give people.

(Naresh Keswani at 00:44:27) That doesn't mean you go and step on other shoes, but you really have to empathize with your customer and really understand the problem statement. And if that means you have to own the problem, not the role, I think that was the defining moment at least for me. When I started to think broadly that I need to solve this problem, and then I need to solve it myself or convince others that that is their problem as well. Even if they work for me or they don't, that's the only way we can get to success.

(Joel Beasley at 00:45:03) Yeah. I'm gonna ask you the same question, Yev, unless if you want a different one.

(Yev (Yevgeny) at 00:45:08) No. It's a good one. I'm gonna be a little simpler. When I started my career, and it was in retail, I've always been in retail. I was 22 years old working in quality assurance, QA, for a company in retail technology. Was there for about maybe six months doing normal tasks, boring, frankly, in QA until I was asked to go to the client. If I remember correct, it was in Australia to help them opening up a store with our solution. That was my first trip to the client, first time I was there. Seeing the effect of your product in the field, seeing how stressful it is, how impactful it is when it's serving the clients, and how stressful it is on the person that implementing it changed completely my way of looking at things.

(Yev (Yevgeny) at 00:46:15) Probably made up my career and everything else and shaped up how I work. Always start from the problem, understand that other side, understand why and what is done, and then work backward to implement what is needed to succeed in this goal. I'm following that advice to myself overall all of my life.

(Joel Beasley at 00:46:35) Amazing. Thank you so much for sharing that. And, Naresh, Yev, this has been absolutely fantastic. We made a podcast. How do you guys feel?

(Yev (Yevgeny) at 00:46:43) It's good. It's great.

(Joel Beasley at 00:46:45) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.