 Hello, and welcome to theCUBE and our super cloud for event on GenAI and more. I'm Rob Streche, managing analyst with theCUBE Research. And today I'm super excited to have David Glick, SVP Enterprise Business Services for Walmart joining us. It's really, you know, again, getting it from the front lines of people who are applying GenAI to the problems they're solving for their associates. So welcome on board, David. Thanks for having me. Yeah, it's excited. Again, I'm a happy Walmart customer, by the way. We have one here in Westboro, Mass that I do frequent. So really appreciate, you know, all the work you do. I think the apps and being able to get there and how you're going to apply GenAI to that customer experience and for your associates is really exciting. But let's kind of go into one of the announcements that you had just a month back where, you know, Walmart shipped GenAI so that the associates could really safely use GenAI features without worrying about, you know, IP leakage and security concerns. Obviously IP leakage is a huge concern for most companies. They're really worried about public commercial models that are tuning themselves and getting smarter based on the prompts and the information put into prompts. So why don't you first kind of give us an idea of Walmart's view on build your own versus buying GenAI? Yeah, well, first let me say that Doug McMillan has come out publicly and said, we are going to use GenAI in all sorts of ways in our, across the enterprise. You know, my job is to take care of associates and they'll take care of customers. And so the way we've organized is that we have a team, a platform team who is making sure do we have the right firewall set up? Are we using the right models? Are we doing it cost effectively? And then I'm an application team. And so I have the fun part which is taking care of our associates. And, you know, one of the things we did even before GenAI is we said, you know, the, we want a consumer great experience for our associates. So they don't have to go to a third party after approve us expense reports, another app to go order food at the cafeteria and another app and so on and so on. How do we bring this together in a consumer grade application for our associates? And we call that me at campus. And so they should be able to do everything they need to do from an HR and finance and enterprise perspective in that app. As part of that, we shipped last month my assistant. And that is gives them the ability to safely put in whatever experimentation they want to do with GenAI. So it might be, I want to write a press release. Can you write a first draft for me? I would like to create a product detail page. We don't even know the many things that our associates will come up with. And so one of the things we want to do is basically crowdsource ideas for the next killer app for GenAI. And by allowing the associates to ask their questions, we can monitor that and figure out what people are using it for. Yeah, so you're doing the prompt engineering part to get a better idea for what they're doing and how they want to consume the data and what questions they have that kind of leads to the whole, how did you get here and where did the data come from? How did you train the model? Did what base model were you using? And how did you really jump into this deep end? Because I think that to me, it's the getting started part that a lot of companies are sitting there, scratching their head going, okay, we have a use case. And you talk about the one with the associates and I love the one around press release. I think that's a perfect example of running a press release has some structure to it and understanding, okay, you're missing these pieces or you have all of the data. How do I summarize that? How did you really start, get started and how did the model develop, I guess? And how did you train that model? Yeah, they say, how do you eat an elephant one bite at a time? And so the first thing is to take the first bite. And so we wanna two different ways. First of all, as I said, we have, we use many different models from different vendors and open source. And the value of the basic model like open AI bar, everybody's done a great job of creating the basic model. For us, it's like, how do we put something on top of it that can do things specific for Walmart? And so we came at two directions. One is, as I said, from the bottoms up, let's crowdsource what the killer app is, but also we wanna get started and we don't wanna wait for that. And so we sat down with our team and said, what can we do today? And LOMs are great as chatbots and great at summarizing things. And so I own the benefits help desk. So our associates, when they have a question about, can I open a 401k or am I covered for this procedure? Like they have a, we have a benefits guide, but sometimes it's complicated. And so they call our help desk and the people at our help desk memorize a 300 page benefits guide. And so if you think about which is better to memorize a 300 page document, is it an agent or a computer? And so what we've done is we love where people led. So we love the empathetic interaction and the people interaction that our agents have, but we wanna give them some help or augmentation. And so the first thing we did is put our benefits guide into the VectorDB, right? And said, like let's use whatever LLM, we can switch back and forth and see which one does the best job, but let's put our information on top of those. And that actually went pretty quickly. The magic of Gen AI, like there's good things and bad things about Gen AI, but the magic is you can go fast, right? One of the engineers said, oh, I'm gonna try this over the weekend. And Monday morning, we were able to answer 95% of the top 200 benefits questions, just as good as the agent did with the AI. And so that would have taken us months or years to do, previous to Gen AI. Yeah, no, I think that's the key, right? It's the human in the loop activity. You're making them smarter, making their job easier, actually more enjoyable in many cases. And we see that even with developers and things of that nature, where using their co-pilots of every elk, I guess, out there. And I think that to me really makes a lot of sense is that you're basically taking a base model and you're enhancing it with your Walmart specific information, keeping it secure. And security had to be a big piece of how you approach this and how you went about this. Yeah, security is job one. We want to keep our people's data private. We want to keep our company data private. And so I'd say recruiting is P zero, but security is P negative one. There's no priority above that for us. And so we built the right firewalls and we built the right guardrails around it. So whatever LLM we use, we keep our data inside our own firewall. Yeah, you talked about the My Assist for me at campus. You know, what are some of the other things that it can do and some of the advantages that you've, you know, that associates are starting to see out of it? Yeah, I mean, we've had associates who come in and what we're finding is when we tell them, do this, this and this, that doesn't work that well. We say, just try it. And you could create a product detail page or you could write a press release, you could summarize the document. There's no end to the ideas. And so as we're getting a month's data, two months of data, we're going through and looking at, you know, which ones are the ones that are good ideas? Which ones are the ones that are used frequently and effectively crowdsourcing this rather than having a bunch of senior executives sit in a room trying to figure out what the right thing is. Yeah, I think that's the thing is to your point, you know, how do you find the next killer app for it? You know, on your other side, I've always been saying that one of the things that I would love is me to be able to put in into an LLM, into, you know, the app and I'm walking down the Walmart aisles and I've said, here's the recipe I'm gonna make. Warn me as I walk past the thing like the garlic I couldn't find yesterday when I was going in to go and find it. And then, so I don't have to double back three times to figure it out because it moved from one shelf to the other. I think things like that. So are you, are you starting to see where the associates are coming up with things and you're being able to tune the prompts and things of that nature? Yeah, I mean, I love your use case of, you know, making it easier for customers. You know, we want to make it easier for associates and usually they make it easier for customers. You know, Walmart has launched over the last several years the ability that have products delivered to your house. So you don't even have to go to the store or you can go to the store and pick it up and that's quicker or, you know, the next gen, the next, what is it? The next frontier is like, how do we know where you are in the store and we can guide you through the store? And that's a great application for something that makes customers convenient, makes things convenient for customers. Whether we use Gen AI or whatever technology what we really want is delightful experiences for our customers and delightful experiences for our associates. Yeah, I think especially the associates part, I think again, you're looking at it and helping people have a better work life balance and getting their job and more job satisfaction. I think to me that totally makes sense with that. And, you know, what are some of the things that you're seeing that could be potentially, you know, that next killer app for Walmart? You know, I love like, I think about, you know, how do we answer questions for associates? How do we not have to have them ask questions, right? You know, the best question is the one that never gets asked because we make it simple for them to understand all of the complexity that it's around them. And, you know, as technologists and developers we need to take complex things and make them simple for the users. And, you know, whether it's calling the help desk or understanding their benefits or understanding their financials, all of these things are things that we want to do to help our associates live better lives. Yeah, no, that is, I think the number one thing a company can do is really look out for their associates. And I think it's fantastic to hear that really Walmart's investing in that. And I think that to me is one of the things where this is really, you know, people are always worried and scared about Gen AI. Have you had to do some training or to get people to embrace this and understand that, hey, it's not only safe, it's we've developed it and here's how it's gonna make your life that much easier or has it just been organic that they've been able to pick it up and run with it? You know, Walmart's a people led, tech powered and how many channel retailer we say in our statement. And so we don't lose the part about people led. I mean, one of the things that's making adoption easier is that Doug has come out publicly, both internally and externally and said, this is important. We are a big tech company and we are going to be on the front edge of this thing. And so it turns out that driving adoption when your CEO is completely bought in makes it simpler. But we are also, we have a program, we do what we call idea jams. And we sit down with the business folks and the technologists and say, what would you like to do with this? And spend two hours like explaining to them, you know, what are the capabilities of LLMs? What are Walmart specific capabilities? And then you come up as the user, tell us what would make it easier. And I think you mentioned earlier, like, could we create 10 case or could we create these documents which are pretty formulaic? They have the same thing in each quarter and we have the data and we can now with the LLMs we can write what we want to with the narrative. And so, you know, whether it's finance or legal or people, everybody's coming up with these applications and it's fun, it's great to have a full funnel for a technologist, right? We have no shortage of ideas and now we can go start knocking them out. Are you seeing that you're applying Gen AI as well and actually the development cycles as well? Are you getting into that as well? That is on the docket, a little bit outside of my space but yes, we believe that there is immense opportunity to be unlocked there. You know, we want to be very careful, you know, our code base is the company jewels. So we want to be very careful, even more careful than some other things of keeping that inside our firewall and making sure we do it safely. No, that totally makes sense. I think that it's such an exciting space and I think that when you start to look at everything that can be done, you know, leaning into it, as you said, having your CEO really stepping out there and saying, hey, this is really important to us. We're really a technology company, technology enabled but people led. I think, I love that. You can't take the human out of it because I think there's certain things that just, yeah, it's great. Like you said, the 300 page doc or writing a 10K for the CFO's organization where it's very formulaic. The answers don't change very, maybe they change year to year. And in fact, you can see, you know, an LLM is actually good at giving you that for like health benefits and stuff like that or how they've changed because how many times have you gone through and tried to figure out, okay, what really changed in this health benefit sheet? And I think to me, that's a perfect use of an LLM. Yeah, and, you know, reading, interpreting and then spitting out the answer rather than spending hours working on, you know, trying to figure out these things which actually aren't your job, it's just distractions you know, one of the things that I found is like everybody wants to do a good job, right? If you give someone a choice, they'd rather be effective than not effective. And so if we can give them the tools to do their jobs really well, like I'm not a lawyer but I can imagine summarizing a 500 page brief like is it the most exciting thing? But if I can get a summary and then do analysis on it that's using my degree. And so we're looking for ways to make our associates more productive and making their jobs more fun. Yeah, I think that's a killer way to look at it is making people more productive. I think that, you know, again, we're big fans of Gen AI and AI in general here on theCUBE and I think, you know, again, that's just the way it should be. Any last thoughts on this as we come to the end here? My last thought is watch out for Walmart. We are big tech, you know, maybe you don't think of us in the same breath as Google, Microsoft, Amazon, you know, the West Coast companies, but, you know, we've got a big engineering team. We've got 25,000 engineers focused on making the lives of our customers better and making the lives of our associates better. And Gen AI is a great tool which is gonna allow us to do this really well. Yeah, if somebody doesn't think of you guys as a tech company, they've been living under a rock for the last like 10 years. I go back to having worked with Walmart even in the open stack days back, you know, when that was a thing before Cloud really had taken off and how was build your own Cloud. So I appreciate you, David, coming on and really sharing what Walmart's doing with Gen AI. You know, again, this is really super exciting stuff what you're doing with my assist and with me at campus and helping it really be smarter to help your associates. Well, thank you for having me. It was wonderful to be here. Okay, thank you. And remember, you can stay up to date with all things AI by visiting gen angle.com. Thank you for watching theCUBE, the leader in high tech enterprise analysis and coverage. Stay tuned for more with SuperCloud 4.