Morning. Good morning. Good, good 8:00 A.M. morning. I'm Joe Teklits with ICR. Welcome to day two of the ICR Conference. If you're on the webcast, welcome to the ICR Conference. With us this morning for a keynote presentation or fireside chat is Daniel Danker, Executive Vice President of AI Acceleration Product and Design. In his role, Daniel leads tech product development using AI, including integrated AI-based tools and solutions in Walmart's omni-retail businesses to drive faster growth. He partners with the business to identify ways to grow using AI and generative AI-based technologies, leveraging strategic partnerships with leading tech companies, which we'll talk about. This includes customer and associate-facing solutions, as well as AI products in the company's advertising, data, and commerce businesses. Welcome.
Thank you. It's good to be here. It's nice to see you all.
Great. You were at CES last week, NRF over the weekend, ICR today.
It's been a busy week. Most weeks aren't like this.
Okay. Well, we appreciate you flying. It's one of those mornings where you woke up and you didn't know where you were, probably.
It was, indeed.
My first question, just going to get to your background real quickly. We all talk, a lot of us have kids at that college age going or leaving, and you tell them, it doesn't matter what you major in, you just go to school, study hard, and you'll figure it out when you get out. So for Daniel, I'm curious how a Mass Comm major from Berkeley ends up running AI for Walmart. How did that happen?
That's a great question. Gosh, careers are not a straight line, are they? It's quite non-linear. And I had a whole bunch of experiences along the way that really I felt kind of prepared me for this role. But I grew up in the Bay Area, near San Francisco. I think it runs in the water there.
It must.
I mean, I think we're all just absorbing all of the technology changes that are happening and kind of seeing those changes happen in real time, so I'm not sure that what I majored in in college ended up having as much impact as perhaps the water I was drinking, but it really, it's truly in the environment, and it's been a gift to see it all unfold.
Well, that's fantastic. And we had a list of questions, and we're standing in the back of the room for 10 minutes talking about all the changes that have happened to you in the past week. And then Simeon Siegel walks in, and he starts asking some questions. So now our list of questions is basically on fire, and we're just going to wing it. But let's just start with this. This is a consumer-oriented audience. It's not a tech audience, so we're going to keep this kind of at that level. At least we're going to do that for me. Start with just generative AI and agentic AI. We were talking about generative AI a year or two ago. Now we're talking mostly about agentic AI. And agentic commerce, just define those for us and the differences and what they mean to you.
That's a good question because the definitions keep changing. I think that it's almost worth rewinding the clock just a quick moment, and you said it's a consumer-oriented audience, not a technical audience, but I think we should test those limits for just a moment here. For a long time, we've been talking about things like personalization, a familiar term. We've all talked about it, and that was always built on a technology called machine learning, and machine learning is just a fancy way of saying pattern matching. The computer notices certain patterns, and it assumes that they'll repeat themselves, and that takes you pretty far. Like if you tend to buy the same milk every week, then when you search for milk, machine learning will recommend the same brand of milk and the same size and whatnot that you normally buy, so that's helpful.
But it stops there at pattern recognition. And so let's talk about some of the pitfalls of pattern recognition. The week after Thanksgiving, machine learning will recommend turkeys. Why? Because a whole bunch of people bought turkeys the week before. Kind of missed the moment, right? AI is different. AI actually interprets. It can form a level of understanding. So it's actually built on machine learning, but it forms a level of understanding. AI will recommend turkeys the week before Thanksgiving. AI will notice that you're buying milk, and you probably also, by the time you add the milk and the flour and the eggs, it's like you're probably making pancakes, and it'll recommend the other things you need to add. It understands quite a bit better. And that understanding is a step change in technology.
In my mind, it's a step change in the commerce experiences that we'll be able to deliver with that technology. So that's AI. Now you took it a step further. You asked about agentic. Agentic AI just means taking that level of understanding and starting to take action on behalf of the customer. So you understand your customers so well that you might just be able to do certain things automatically. Things like if we understand that our customers, we understand them so well that we know that they're probably running out of laundry detergent, which by the way, is not as simple as it sounds. It's not just noticing how often they buy laundry detergent. It's also recognizing that because they tend to buy a gallon of milk, they're probably in a four-person household. So they're probably going through laundry detergent at a certain pace.
And so we know the right moment to recommend that laundry detergent. Agentic AI will then just go ahead and send you the laundry detergent before you even run out of it. So that's the promise. We're on a journey toward it. We're not quite there yet, but we're moving fast. So I think it's coming.
Do you know that the customer wants it? I mean, there's tech companies that push their technology, runs through retailers. We all think it's great. Does the customer even want that, or they like the old-fashioned way of shopping?
It's a great question because I think that over the last year, year plus, we've all been playing with the technology to understand it better. So a lot of things have been built that won't necessarily work. A lot of things have been built that don't necessarily reflect exactly what the customer wants. I try to simplify these things. I think of us all as carpenters, and we've all been using screwdrivers, and suddenly someone showed up with a drill. And the first thing everybody assumes is that the drill will make it so you can build things faster. And that's true. But it also means that you can build things bigger. There are things you couldn't build without power tools that you can with power tools. AI is a power tool for us. Now, for the last year or two, we've been tinkering with it.
We've been trying to understand this new drill and see what it's capable of. Didn't really know what we wanted to do with it yet, but we were kind of exploring. This is the year where tinkering becomes transformation. This is the year where we've built a level of mastery around that, and we'll start building things that deeply address customer problems. But I'll give you a few examples of customer problems. And let's just see in the room if you feel like these are relevant or not. An example of something that you can do with AI that you genuinely couldn't do before. So for 25 years, we've been using e-commerce, and we've been searching for clothes, and we've been scrolling through long lists of photos of other people wearing those clothes. Sound familiar?
And we're doing all these mental gymnastics while we're scrolling to imagine ourselves wearing those clothes. AI will just show you wearing those clothes. You're not going to scroll through a list of other people wearing them. You'll scroll through lists of yourself wearing them. You might not even scroll through lists. You might take a picture of something in your closet and say, what would go well with this? And it'll just show you. And it'll show you with yourself wearing it. And that, I think, is a simple example of something we would all want. And a simple example of something we can all understand would drive way more commerce and way better commerce than we've ever been able to do. But that's not a new idea. But we haven't had the drills. We didn't have the technology with which to do it. Now we do.
What is that time frame? How big does that become, and how quickly does it happen? And I'm sure there's no end goal. It's just going to be continuous, continuous. But is there a date where we're really going to be using everything you just described?
If you go back and look at technology transformations, even though you get leapfrogs and step changes in the technology, you've probably noticed that the actual change happens gradually in a whole bunch of places. So every little step in the process changes and gets better. And you kind of don't notice it as it's happening, but if you look back a year or two later, you see a big change. That's what's going to happen here. It's not going to happen overnight. It's also not going to happen all at once. But I do think this is a year where we will be delivering transformative experiences in commerce. So you will definitely look back a year from now and say, "Gosh, that's quite different than how I used to use Walmart. That's quite different than the expectations I used to bring when I opened an app."
Walmart seems to be very much at the forefront of utilizing all of these technologies. That's not table stakes, but at some point, if you do it, it's going to be table stakes for everyone else. Is there a risk to being too far out front?
The risk is that we build a few things that don't stick. I'd say there's a much bigger risk to not being out front. And that's why I like to use the analogy of the drill. Can you imagine if you hired a carpenter and they showed up with a bag full of screwdrivers and no drill? You'd think that's crazy. And you probably wouldn't hire them again. And I think that this is that step change. And we're going to lead because we think that it can do things for customers that we genuinely couldn't do before. Not every single thing we try is going to work, but the only way we'll get to the thing that works really well is by trying a lot of things along the way. And we're very lucky to be able to do that.
It's something that I really value about what we can build.
So on the other side of every risk is reward. And so is that first- mover advantage, customer acquisition tool? Is this something that's going to drive your business and kind of widen the gap between you and the competitors? Is that why you take the risk?
We're generally, a little bit more simplistically, just driven by our customers and what we're trying to achieve for them. We're really customer-obsessed in how we approach these conversations. And we start from a list of opportunities and customer problems that we're trying to resolve. By the way, when I say customer, I mean the broader customer. We're talking a lot about the consumer experience, but this applies equally to technology that involves our supply chain and how we get goods to customers. It also involves technology that we're putting in the hands of associates that are in the store serving customers. And by empowering them with more AI tools, which we now do, they're able to serve customers in much, much better ways.
They now have access to a wealth of information about the products that are on the shelves, about what's coming, about where things are in the store, and they can help customers with all this information right at their fingertips, so the customer is a broad set of customer groups, and I think I don't view it as a risk to experiment. I view a much bigger risk to not experimenting and not figuring out how we solve each part of the value chain there.
Right. Okay. Let's quickly talk about your partners, specifically OpenAI and ChatGPT, which you announced that partnership a few months ago. And then Gemini, which is Google, which you announced this Sunday, this weekend. Talk about kind of the differences, what's happened with OpenAI, and now why Gemini and what the differences in those two platforms will be.
We view a big part of our role as finding customers wherever they are. We need to reach our customers, whether they came directly to us or whether they started their journey elsewhere. One of the big reasons for that is that folks tend to come directly to Walmart when they have what we often call commercial intent, meaning they want to buy something. They know they want to buy something. They come to us for that. There are a lot of journeys that don't begin with commerce, but often end with commerce. One of my favorite examples of this is if there's a wine stain that you're trying to get rid of or get out of your carpet, you're very likely to go to a product like ChatGPT or Gemini and say, how do I get this red wine out of my carpet?
You weren't thinking about buying anything in that moment. You were genuinely trying to get advice. Somewhere along that journey, ChatGPT or Gemini might say to you, there's this product for the kind of carpet you have and the kind of stain you have, and you can press one button and end up shopping that on Walmart. These journeys that begin with something that doesn't look like shopping, but actually end with shopping. We really want to be a part of those journeys as well. We view this as a huge growth opportunity because it enables us to reach customers in those moments too. That's a big part of the why behind those partnerships. Now, it's very early days in terms of how those integrations work.
And one of the things that we're really excited about in what we announced this weekend with Google is that we're essentially having their AI agent, Gemini, partner with our AI agent to create a unified shopping journey. And that's a fancy way of saying that when a customer discovers something on Gemini, Gemini might recommend a new TV, a wine stain remover, et cetera. It calls up our agent and enables us to offer the customer a very familiar and personalized experience directly within Gemini. So almost imagine it like a window inside of Gemini where our shopping agent kicks in and helps you complete that purchase. Now, why is that important? Well, very rarely do we find people buying one item. Quite often, you buy one thing and you're on ramp to buying a full basket of goods. And we're really good at that.
We offer a tapestry of products that customers want, and so that's one element of it. Another element of it is that for the most part, our customers aren't just customers. They're often members, and so they're getting great delivery fees and a great experience that's really attuned to them and has gotten to know them over quite some time, and so that member experience shows up directly within Gemini, which is pretty cool, and then the last one, and this is almost mechanical, but I think it's actually quite reflective of how people shop. What we see is that people add things to their Walmart cart throughout the week. They don't necessarily check out right away, but you realize you ran out of dishwasher pods. You go in and you add it to the list, and so the basket is built over quite a few days.
What happens is that that basket now automatically joins the item that you discovered on Gemini, and so when you order, you receive one box with all of those things together, and that might sound simple, but people really do have this kind of thing in the back of their minds constantly. How many additional boxes and orders am I generating with each move? And so having that discovery experience start inside of a Gemini context, but end fully inside of your Walmart relationship works really well for customers, and it's something we're really excited about, so an important early path on this kind of integration, figuring out how do these surfaces that happen off-platform join up with all the goodness that we can offer to customers when they come straight to us.
Fantastic. OpenAI is, that sounds like 2.0. That sounds like agentic commerce 2.0 if OpenAI or ChatGPT was 1.0. Is ChatGPT going to leapfrog and go to 3.0? Or what are the next iterations of that? It seems like these two agents are competing against each other now in a way.
Yeah, we view our role as working with both of them and others, indeed, to figure out how these journeys should work, and so I think you're going to see a constant evolution on all of these products, and maybe over time they diverge or maybe not. It's really hard to predict. We view our contribution to it as being this really important combination for customers of massive assortment at a great price, predictably. Customers know if it's coming from Walmart, I know I'm going to get a good deal and at great speeds. And that breadth is really important and quite hard to find, and when you imagine that you're operating in an environment like ChatGPT or Gemini, where you could be asking about anything at all, it's not like every journey will involve eggs. It's completely open-ended what you'll discover.
Having that huge assortment plus speed plus price is a really valuable combination for the customer. I think that's true whether it's in Gemini, whether it's in ChatGPT, and however those integrations evolve and those products evolve. I think that that fundamental combination of assortment, speed, and price reflects a really common need across all of them, and that's how we view it.
So in 1.0, I would go to ChatGPT, and they'd give me a few options of where I could buy something if I was prompting them for the best pancake mix or whatever that would be. And then I would have to click through it and go to that retailer's website, like your website. 2.0, Gemini, it's all happening within Gemini, right? But it's still a Walmart interface within Gemini.
That's right.
Will there be a time when I just say, "I want this, please send it to me," and I don't have to do anything else? It shows up at my door the next day.
I think these user journeys are going to become simpler and simpler and simpler over time, and there are times when a product you discover is truly an on-ramp to other things, and there are times where you just get it and be done, and we want to serve both of those needs.
How do they choose Walmart to fulfill that purchase versus all of your competitors? Or who chooses that?
So it depends on the environment that you're in. But if you are inside of ChatGPT or Gemini agent, of course, they have an algorithm that is determining what products they're going to show. But if you kind of go back to what would make for a successful algorithm versus an unsuccessful one, a successful algorithm is going to be one that serves the customer need. If you're going to let a computer start making decisions for you, it needs to make decisions that are similar to the ones you would have made yourself. And that's why I believe that the currency, the most important currency in an agentic shopping world, is actually trust and affordability. I think without trust and affordability, it's very difficult for customers to hand the wheel to someone else and expect that the right thing will happen.
So it's perhaps more than just convenient that those are really core to Walmart's values. It's been core to our brand and to our history and kind of how we view ourselves as serving customers. But I think those values are translating to an agentic shopping world in a way that maybe we wouldn't have even been able to anticipate, but kind of gets at a core human need. And so I have high confidence that we show up well in these agentic shopping experiences, even the ones that don't start on our own app, because those core customer needs, they've been true as long as time, and I think they will continue to be.
So, I do want to ask about price. And you kind of went there, so I'm going to go there. For the customer over time, what does this do for prices? Does that agent search for the lowest price and this helps customers more easily find the lowest price? Or is that if that's what's important for the customer, or is maybe that's not important for the customer?
I think that's where you kind of hit it on the head there. I think that some customers are very value-oriented, and they are going to orient themselves toward the items that have the lowest prices. Some customers might be less price-sensitive. By the way, more realistically, all customers are price-aware on certain products and less price-sensitive on other products. And this is why it all brings us full circle back to the beginning of this conversation around personalization. If you really, truly are going to understand your customer, then you'll understand that they like to save money on paper towels, but that they like to splurge a little bit more on their produce or on other things. And that's been part of Walmart's assortment strategy for a long time, to be able to serve those customers regardless of where they are kind of on that thought process.
This is why it's so important for us to be able to offer that assortment, but also match it with our understanding of that individual customer.
Okay. So that gets us to data, right? You have, with your customer in your stores, on your website, you have a lot of customer data already. Gemini, ChatGPT doesn't have that data. So it doesn't know as a customer what I prefer. Are they relying on you for that data? Is there a partnership there? Are you willing to share that data? Or do they want you to come, they want to come to you to fulfill everything, and then you own that data?
OpenAI and Google will personalize based on the things that are done on their inside of their products. The experiences that customers have and the data that we build up on our own products and in our own stores, indeed, stays within Walmart. And we do not share that with any other platform with OpenAI or Google. There are some small bits of data that go back and forth in order to complete a transaction that you can imagine you would need to know. You would expect Gemini to tell you, okay, your order's on the way now. So there are little bits of information that need to go back and forth for that. But even for that, we're extremely clear with the customer that some information would be going back- and- forth. And they have the option, they have to explicitly say yes before we move any data around.
So we're pretty protective of that.
Okay. I'm going to stop going down this rabbit hole. We'll pivot to something else, but I wrote the word disintermediation down because I think that's on some investors' and analysts' minds. Are you, five years from now, this turns out to be a positive? And how is that going to show up as a positive? Or is it disintermediation where that customer that used to go to walmart.com is now going to Gemini and all of a sudden you're losing a transaction here and there because of Gemini? How do you view that kind of risk-reward?
I see this very clearly as a growth opportunity. But let me explain why. There's a few key components that make this a growth opportunity that we've been very intentional about and which I think set us up really well to serve customers even as the environments shift. One, it goes back to that combination of assortment, price, and speed. That combination means that Walmart shows up a lot inside of Gemini and ChatGPT because we offer such a complete package for customers that doesn't just serve one need, but serves a whole bunch of needs. Two, the approach of having their agents work together with our agent and creating an experience, a Walmart-powered experience that shows up directly inside of those environments means that we're orchestrating an intelligent handoff between the products, so rather than being invisible on the customer's journey, so we're actually not receiving just orders.
We're actually receiving customers who are in the midst of placing an order, and we can take those customers on a journey by offering additional products, connecting it with the rest of their cart, enabling them to benefit more from their memberships, et cetera, so that connection between the agents is extremely important. And then finally, as I started with, this is not really about taking customers that come to us because they know they want to shop and giving them a different experience. This is about recognizing that there's so many shopping occasions that don't begin as a shopping occasion, and those are moments that we want to be able to serve, and this enables us to serve those moments better than we've been able to do before.
Great. I want to go down the income statement real quick because that's how this audience would think, right? We talked about the revenue opportunity. What happens to the gross margin, quote, opportunity if there is some sort of sharing between the agent and Walmart that maybe wasn't there before because customers are gravitating to using this tool more and more? You feel this is a positive too? Are you sharing revenue now? Is this positive or is this a risk to the margin?
I mean, we've worked with other platforms for many years in order to show up in the moments where customers are searching for products and whatnot. This isn't actually fundamentally different from a search ad or any other engagement where we work with partners and platforms that customers go to to make sure that we show up when they have a need that we can serve, so in that sense, it's not fundamentally different, and it comes down to the same measurements that we would have done years ago in a more traditional search world, which is assessing whether this is incremental, whether this is causing us to, are we super serving the customers that are already coming to us and it doesn't make much difference, or are we reaching customers in moments that they wouldn't have necessarily thought to come to us?
And that's the part that gets me most excited. And I think that's where it's going to go.
Back to the revenue part real quick. I do want to ask, what are you finding right now is the kind of highest use case? Is this being used for kind of high- consideration product categories? Is there a differentiation in demographics of who's using these tools? What type of product consideration? Are you seeing any differences there and maybe where that's going?
Yeah, that's a great question. I don't know yet on the demographics. It's a little early to know. But when it comes to what people are buying, this is pretty interesting. You could imagine that this would start with the things that people buy most frequently. That was one hypothesis. Maybe this will go that route. You could also imagine that this would go down the path of the things people buy not as frequently, least frequently, but which require more exploration. And it has very much gone down that second path. So in other words, when people want to buy their essentials, their weekly grocery shop and things around that, they know how to do that. They have tried and true ways of doing that. They come straight to us. By the way, we're going to evolve those experiences quite a bit over time, which we can get to.
But those are well-understood journeys for customers. The part that's really exciting about this is that it's unlocking an ability to serve customers who aren't totally sure what they need. Maybe they're trying to shop for a TV that'll fit. Maybe they don't know how big of a TV to put on a wall of that size. Maybe they aren't sure if it'll work with their gaming device or whatnot. And so that kind of interrogation and conversation really takes a few back- and- forths before they know exactly what they want to buy. And so it's generating those kinds of orders.
I think fashion is going to be another category where we're going to see a lot of change there because it is an area where there's a little bit more room for inspiration and a little bit more room for seeing yourself in it and seeing the user journey change quite a bit from trying to figure out what to put in a search box and then scrolling. So we're going to see a lot of that. I think baby is going to be another category where we're going to see a lot of change. New parents, it's been a few years. It's been four years now. But I remember when our youngest daughter was born, and gosh, we had so many questions about what's the right thing to buy. How do you do this? How do you solve for this?
They, for some reason, all happen at 2 in the morning when the baby's crying, and you have with one hand, you're trying to figure out how to solve this thing that you're unfamiliar with. Those are the kinds of journeys and customer experiences that I think are going to benefit most here.
Fresh food's currently not part of this.
I think fresh food, where we're going to see a change, is that the things that are on repeat will happen automatically. That the agentic experience will take essentials and just make it so you don't have to do the same thing week over week, finding and buying the same product you buy every week. I think that burden will get just completely wiped away by the agentic experiences we're building. A lot of those are going to be the agents that we're building directly into the Walmart app. I think that these additional categories around baby, beauty, fashion, auto care, there's a whole slew of categories, electronics, et cetera, we're going to see a bigger transformation because the customer will be able to have more of a conversation with us about what it is that they're looking for.
Okay. Go to the other end of the spectrum. Stores, they still are going to exist in 10 years, right? What are they going to look like? So how does this experience change the store experience 10 years from now? Look out into the distance. What are we going to be doing with stores?
First and foremost, people love to shop. That's not changing. People love to shop. And so we are, what, 25 years into e-commerce, 15 years into grocery delivery, and still 20% plus or less than 20% of all shopping is happening online. People go to the stores. And it's because you get to touch and feel, and it's a delight for the senses. You walk in, you get this panoramic view of thousands of products that have been very intentionally placed to help you shop and discover. So it's an enjoyable experience. So I think it's actually a mistake that I'm seeing a lot is folks thinking you either have the store or you have online shopping powered by AI. I think actually you have shopping powered by AI. And our goal is to digitize the in-store experience to the same extent that we do the online experience.
There are two different modalities. In the store, there are so many opportunities to help customers with a lot of the same decisions we just talked about online just happening physically in the store. That exact same story we just started with around fashion, where you could see yourself wearing the clothes and where the AI agent can help you know what to pair with what applies just as much in the store. I would use that all day long. There's a lot of categories where that's the case. We're excited to be building a very omnichannel experience that spans the store and online. I think you said 10 years. I'm not even sure we need to wait that long. Our customers already open our apps in the store, both on the Walmart and Sam's Club side.
And they're already using our own apps in the store to shop better in the store. So I think we're going to see a big change there. And I think it's a pretty exciting one.
And you mentioned the app. What does the app look like 10 years from now or five years from now if you want to pull it in a little bit?
Yeah, well, I think we're going to be doing less scrolling. We're doing a lot of scrolling today. We've been scrolling for a while, and I think that as our products understand our customers better, and instead of showing you eight different tomato pastes when you search for tomato paste, we'll show you the tomato paste that we think you want, and we'll also show you the other things we think you need. Because by the time you've added the tomato paste and the ground beef and the mozzarella, we're pretty sure you're making lasagna, and we don't need you to search eight times and scroll through many, many pages just to add the basil and the tomato sauce and the ricotta. I make a lot of lasagna, and so we're going to be doing a lot less scrolling. Those experiences are going to become more human, more connected.
They'll understand your intent, and they'll serve it up to you much more easily.
Super. Sticking with the app and e-commerce, Sparky.
Yeah.
How's Sparky doing?
Sparky is just, he's so happy, so happy. Sparky is our little happy face at the bottom of the app. I said he, but we actually don't know if it's a he or a she. So we should leave it at that, actually. And when you invoke Sparky, it comes up, and it's a chat interface that lets you interact with us in a slightly different way. Now, interestingly, today, you have a search box where you type things like tomato paste. And you have Sparky, where you type things like, how do I get this wine stain out of my carpet? And I do think over time, they probably come together. And I think that will make for an even better experience. Today, Sparky is doing a number of things really well. First and foremost, customer service.
If you can't find your order or if you need to return something or if something was broken or if you just need another of something, you ordered one, you meant to order two, Sparky's great at that. And customers are naturally going to Sparky to do that. And I think it's because it isn't totally obvious where else to go for those kinds of needs. They're a little ambiguous. And so a chat interface is really good for those. Some of the other things that Sparky's doing well, as of last month, Sparky can wake up on its own and just say, "Hey, you open the app," and it notices that you tend to buy many of the same items week over week. And it'll just say, "Do you want to just are you just here to buy these same items? We'll get you started with that."
So it saves a whole bunch of time in what would have been a lot of searches and a lot of time spent. So it's early days, but it's off to a pretty good start.
But you have no plans on Sparky being what Gemini is today?
Sparky is powered by an LLM in the same way that a Gemini or a ChatGPT is. So you can ask Sparky a lot of those questions that you would ask a more open-ended LLM. And it will help you with those. But we have to start with what the customer's thinking. And the customer that opens Sparky has opened an app that they use to shop. And so we know that their headspace is more in the commerce space. And so everything about how we design Sparky is oriented toward making those journeys easier. We aren't just guessing at what the customer came to do. We don't necessarily know the question that they're going to ask. But we know they're there to shop. We know they're there to do something related to Walmart. And we're there to help with it.
Now, notably, I say shop, but actually, we have a pharmacy. We have an optical department. We have an auto care center. So a lot of the time, I expect customers aren't just going to be coming to buy items. I think they're also going to be engaging in some of the services that we offer in our stores and in our clubs. And Sparky will be able to help you with that. It'll be able to say, "Hey, I noticed you have an auto care appointment, and you have a prescription to pick up. Make sure you get them at the same time. Don't forget one or the other. And you can get your prescription while your car's being serviced," things like that. So you're going to see a lot more of those kinds of journeys show up inside of Sparky as well.
OK. I think we just real quickly should back up and touch on how you've used AI agents internally in the organization as well. Talked about that for a couple of years. But I think it's an important conversation and all is going to blend together at some point, as all of this is. So maybe you can update us on how you used AI internally in the organization, kind of some real success stories. And then you touched on the stores a little bit, but a little deeper into how you're using it, how your associates are using not Sparky, but what's the name of the associates?
Squiggly.
Squiggly. Thank you. How they're using Squiggly in stores?
Yeah, absolutely. It's been deployed so heavily throughout our supply chain that it's actually really important to walk you through. So first, let's talk about the supply chain. Our fulfillment centers and our distribution centers have heavily deployed AI and robotics to help with the movement of goods and getting goods to customers. And I'll give you just one example of why that's so important. If we can predict which products are going to be needed in which stores or to which customers early, then we can push those products down through the supply chain so that by the time the customer is placing an order, it seems impossibly fast. How is it possible that Walmart had that item so close to me at that moment? And it's because we use technology and AI to anticipate that long before the customer even placed the order.
It's easy to focus on what the Walmart app will do with the customer in that moment. Actually, it's only possible because of the use of AI all the way back through the supply chain. That's one example. I think it's a very interesting one. You talked about the associates as well. Next time you walk into a Walmart or a Sam's Club, watch the associates. Watch what they're doing. You'll notice, if you look over their shoulders, that they're all using an app built specifically for the associates that is entirely AI powered. That app is doing things like helping them know which shelves need replenishing first. Super important. You can have two different shelves that are running out of stock at the same time. One of those is a fast-moving item. Another one is a slower-moving item.
You should replenish the fast-moving one first. You'll also notice that their path through the store as they're replenishing is so efficient. You might take a page out of their books. You'll notice they shop faster than any of us, and they're just doing it in reverse. They're putting things on the shelf, but it's because this app is guiding them through the store because it knows the most efficient path through the store so we can get items on the shelves and to our customers as quickly as possible. You'll also notice something else happen. Don't test this, but if there's suddenly a spill the next aisle over, an associate will drop whatever they're doing to go clean up that spill because that's a safety issue. That needs to be addressed immediately.
And that is because the agent on their app will automatically pop up and say, this needs to be done first. Something urgent has come up. That might even be a customer needs help with something on a different aisle, et cetera. So when we say that we want this technology to help our associates serve customers better, this is what we mean. And the fact that every associate has this incredibly powerful AI agent in their hands, in their pockets throughout their day is a really big part of why we're making that possible.
Great. One more question for you. We've got a few minutes left. I don't want to get you in trouble. I don't want to put you on the spot. But the question is, what aren't we asking or what aren't we thinking? And a question that I would say, what problems will AI solve or address in the next year or two years or three years that you're working on but we haven't thought about yet?
Yeah. At its core, this all really does come back down to customer problems that we all know, we all experience. So for us, AI really needs to have purpose. And our entire strategy and plan with AI is built with a purpose that we think needs to be extremely practical. It doesn't need to be overcomplicated. Now, the technology is complicated. The technology is incredible. The customer problems we need to solve are very practical. And so the whole roadmap is built around those. The things you'll see change if you look back a few years from now at the product are going to be that it truly feels personal, that it actually understands not just that individual shopping journey, but it understands your household, your behaviors, your dietary needs, your health needs, the community in which you live. So it will feel much more personal.
We've been using this word personalization for a little bit too long, so it feels like we've done that already, but I would argue that we've just barely scratched the surface on personalization, and AI is going to take us to a new place there. Second, I think the experiences are going to become so much more immersive that if you go back 25 years, what we essentially did was take the physical store with that panoramic view and try to staple it into a 3.5- in screen, and the theory was that personalization would make up for the lack of field of view and immersiveness. I don't think it really has. I think it's better than not the personalization that we have today, but it doesn't quite live up to the promise. I think it's about to.
I think what we're going to do is instead of trying to translate the store directly to a phone, you're going to see immersive experiences that recognize what you're buying. Fashion will be different from pet food, which will be different from your weekly essentials, which will be different from electronics, et cetera, so it'll be hyper-specialized to the things you're buying and fit much more naturally on the device that you're shopping from. I think the devices you shop from are going to evolve. Could be glasses. Could be other form factors. I'm not overly urgent about that because I think that the phone is really functional, and any new device that shows up needs to actually do better than the phone, which is a really high bar, so it's not just because it's new and novel. It actually has to do better.
But I think we're going to see an evolution of the devices themselves. So truly personal, truly immersive. It will anticipate your needs a lot better. So the things that you do repetitively, that's going to get taken care of for you. We're not going to be in a world where we show up every single week, rock up, and do the exact same things. We're never going to run out of laundry detergent again. So those are the things that I'm looking forward to. And I think they're going to happen gradually, as I said. They're not going to happen all at once. So we might not notice every transformation as a transformation. But if we look in the rearview mirror, we'll realize just how big of a transformation is ahead.
Thank you, Daniel. I think we're all a lot smarter. Daniel flew in and out really quickly for this. So please help me in thanking him for being here.
Thank you.