 Welcome back everyone to theCUBE's live coverage of TerraData, TerraData possible. I'm your host, Rebecca Knight, along with my co-host and analyst Rob Streche. We are joined by Hilary Ashton. She is the Chief Product Officer at TerraData. Thank you so much for coming on. Thank you for having me. This is so exciting. Yeah, and congrats on your keynote earlier. Thank you, thank you. So tell our viewers at home and then for those folks listening, what are some of the unique capabilities that TerraData's architecture offers? Like what makes TerraData special? Yeah, I mean TerraData is pretty unique in the market today. I think, you know, a couple of things. One, we came from a world of scarcity and so we take every single resource, like it's our last possible resource. And when you convey that to the cloud in the way that we have, it means that our enterprise scale price performance beats everybody else in the market. And those who are sort of born in the cloud write thought of things like infinite scale. And when I talk to customers, I say, do you have infinite dollars? Do you have an infinite budget? And it turns out they don't. And so really thinking about how you can cost manage what you're doing in the cloud in a way that really delivers value, but is highly tied to price performance. And so that is one of our really big key differentiators. I think the second area is really around our usage of advanced analytics and analytic capabilities. Now, the Gartner Report gave us the number one ranking in all four analytic use cases, which is pretty impressive from where our customers and the reason for that is that we have brought the analytics to the data and many others bring the data to the analytics, right? And so they're moving data unnecessarily. And so the ability to do over 140 different analytic functions directly in database and we're thinking about things like end path analysis. So you don't know what end path analysis is. It really allows you to understand a customer journey, a digital journey of a customer, maybe on your website, maybe in a multi-channel way and being able to string together that journey to make sure that they're having successful journeys. And so that end path analysis function is baked into the database. So that's another key advantage that we have. And then I think today in my keynote, sorry, my keynote was really focused on innovation. You know, some of the things that we're doing with Gen AI, some of the things that we're doing with large language models, and then some of the things that we're doing that are kind of crazy out there from a compute engine perspective. So excited to share a little bit more with you about that. Yeah, we're going to get into all of that. Yeah, well you brought it up. So let's dive in on Gen AI, right? So what is it? I mean, you have ASC.AI, you have some other things that, and it really being a platform for AI. And like you said, how do you bring the AI to the data for that matter? And I think that there's a lot of talk about people have looked at Teradata in one way and are now seeing it in a different way because the data is in there that they want to use this. So they have a lot of first party data in there. They have a lot of buyer behavior and other types of data in there that they want to understand and use and say, hey, we want to make better recommendations. You're in the store, you just went down this aisle, you tell us you want to make this particular recipe. You walked past the cans, I don't know, mushrooms or something like that. What are you doing to help customers from a Gen AI perspective? And really, how do you see that the Teradata architecture is really innovating in that space? Yeah, so a couple of things with Gen AI. The first is, as you mentioned, we just launched ASC.AI. We're so excited about putting the power of a large language model into the hands of our customers in two ways, really. The first way is unlocking the capabilities that you have in Teradata today in a way that doesn't require you to read documentation or watch a video, just a conversation of how do I use this function? How do I do this thing? Automatically create some code for you and then off to the races you go. Of course, humans are still in charge. I think it's really important to remind everybody that humans are still in charge of what you decide to do with that. But we give you a really great head start to get productive very quickly. The second area is really understanding your data corpus better. And so using the data that you have in a secure way within your own tenant with the same rules and roles and responsibilities around who has access to which data. But being able to ask simple questions that perhaps you were using a BI tool before. So how many customers shopped in store XYZ during the holidays last year? What kind of staffing capacity should I plan for so that people can check out really easily? So really very strategic questions about the business and over time, the large language model that we're able to deliver to our customers will allow them to just get answers and not have to worry so much about how to ask the question. So we're very excited about that. The third thing that I'll just mention that we just launched, I just announced it a few minutes ago, is API integration with OpenAI and Azure OpenAI. And so this is pretty transformational from a predictive analytics perspective. So many of our customers, all of our customers have vast amounts of text data. They have call center data, right? They've got the script from call centers. They have survey data. They have reviews, right? Who doesn't have a review site right now with how you can gold star things. And so making use of that data is tricky. And so with this OpenAI integration through an API, you can take that text data, run it through Azure OpenAI and then you can start to take advantage of the embeddings directly within Teradata Vantage. And so those embeddings are really numeric representations of that text data. And we can start to understand that this group of customers through our K-Mean segmentation, this group of customers is happy. This group of customers is unhappy. This group of customers is ready to buy a new thing. And this group of customers is still having some trouble. And so that segmentation based on that rich language data, I think is really an opportunity to take your predictive analytics to that next level. And really in this competitive analytic market, get ahead of your own competition for our customers. So those are three of the areas where we're looking at large language models. We're looking at embeddings and we're looking at generative AI within the Teradata Vantage platform. So I think just to follow up on that a little bit, it really means that a lot of times when you hear people talking about things like embeddings, they're also going and putting it in a different database and a vector database or something like that. So you're saying that architecturally the platform that Teradata has really is one platform that can help you get to that next level and get to gen AI. That's right, that's right. So you don't need a vector database. You can take those embeddings, you can use them and you can, through our analytic capabilities, understand the distance between those embeddings and you can start to really use those for critical insights into your customers or the other outcomes that you're trying to drive. And to your point, I think not having to move the data or have data distributed all over the place really allows you to leverage our price performance that we talked about, reduce latency, right? And so you can get to value faster and yet we're open and connected. And so if you have data in other places and maybe that data doesn't make sense to move right now through our query grid capabilities, you can get to that data wherever it is. And so we say use the data, don't move it, don't copy it, use the data where it is and use our remote push down processing to make sure that you're respecting data gravity. You're doing the work where the most data is and we believe that gives you the fastest outcomes at the lowest price. What are you seeing in the product groups that organizations that you work with? Teradata is so out there and has been working in artificial intelligence for a long time and now we're really at this inflection point with a lot of corporate America saying, okay, we want in, what do we do? Coming to you and how do you make sure that you are helping them stay engaged and innovative in the same way that your organization at Teradata does? Right. I think that the, a couple of things. One is the generative AI and large language models. Chat GPT has really taken the nerdy geeky part of analytics and brought it into the boardroom, which I am so excited about because it means you can now have a technical and a business conversation. And then if I were to leverage against that technical and business conversation, I think making sure that what you're trying to do has business value, right? There's a lot of cool things you could do but understanding why, what outcome you're trying to get to has never been more important. And in fact, when we look at, and this is not unique to Teradata, when you just look at data science work, it is fraught with failure. It is, there's a really high bar to actually get something into production. And so we take that as a huge challenge and opportunity to help our customers quickly and easily get to value with analytics, with advanced analytics, with gen AI. And so I think that given our, as you say, like our background and our expertise in this area, our customers are really able to benefit from some best practices and from that technology that delivers that sort of end-to-end experience to get AI into production quickly. And so I think that's key because I think it also leads to that not only getting there quickly but there's pieces that have to be there. Like you need to understand the data mappings, you need to have a better understanding of the data. You have some stuff around data features that you're able to do and even partnerships with people like DVT and others that, I mean, the data geeks that I all know all use DVT for doing some of their modeling. So how does Teradata really help in that space and getting the bolts of the data knotted up and singing together? As you can see. I mean, look, it's hard, right? Data is all over the place and getting to outcomes is critical. And so we think of an enterprise feature store, which is really using data that you know is valid and good data and creating reusable assets from that data. And so this is going to sound kind of funny, but if you go to an enterprise and you say, how do you define a customer? And if you asked five or six different people, you might get five or six different answers in terms of the definition of a customer. A customer might be a household, it might be an account, it might be just these kinds of accounts, but we don't really look at those kinds of accounts. And so having a single version of what an account is is a data product, right? That's an example that now everyone, whenever you do anything with an account level at a customer, we all have the same answer, right? And so making that a reusable asset, and that's a very simple example, but I think it's something that makes a lot of sense is that if everything you're building is derived from that known good data product of an account, this enterprise feature store now allows you to move much more quickly instead of you build your version, you build your version, I build my version, we hope we get to the right answer roughly at the same time. And so that reusable aspect of data as a product and then the analytics on top of that data as a product, I think is core sets of the building blocks that let you move faster with data, but in kind of like harmony, if you want to be like kind of weird about the language, but like in harmony in your organization so that you, you know, a lot of our customers who are here today are in highly regulated markets. They don't have the advantage of being able to just be loosey-goosey with how they define some of these things. And so that data lineage, where did you get that data? A core part of our product. In fact, we just made an acquisition with a company called Stemma that is really focused on data cataloging and data lineage. So knowing that your data is a good asset, how you got that data, how you derived that data, then making it into an analytic model that you can now manage with our model ops and you can now understand if that model is performing the way that you want it to perform, the way it should perform. And if it doesn't, we give you an alert that says, hey, that model over there, you might want to have a look at that. It's starting to drift a little bit in terms of its performance. And then ultimately having that closed loop process where you're just constantly getting smarter and better about those analytic outcomes with the data as a product that you have from us. You know, I think that's great. And after this, I'll send you, actually I did a little write up on data products and how they work and how the rise of the data product manager as a matter of fact. And we also talk a lot about data developers and things of that nature. So I think that just makes total sense to how you approach this in more of a product guy as well. So it's like, you know, I look at it from a product perspective, how do you get your minimally lovable product out there to your point, getting everybody to speak off that? I was saying, you know, one of the great things for instance of LLMs is being able to bring people up to speed when you bring them into your organization very, you know, very succinctly. Very quickly. Yeah, I think that's bang on. And I think when we look at our customer base, when we look at enterprises, they've really moved from being, you know, just a bank, just a telco. They're technology companies that are building technology products against data and analytics. And so it's really tooling them to become product developers with our integration with DBT, for example. Right, to be able to encourage that flywheel to move faster and really enable them to have a very strong CICD process that they deliver. I was just talking to one of our customers, they're like, yeah, we're a software as a service company. And they are definitely not in your Tech 500 list. They have bricks and mortar, but they are in essence a technology company and they rely on Teradata to be that building block for their software as a service delivery, which I think is so cool. So how are you at this conference using the proof of concepts and what you're hearing from customers and using it to sort of evangelize the Teradata approach and model about integrating large language models and sort of using the Teradata architecture. Yeah, I mean, I think that the feedback that we're getting from customers is wow. That is literally what, every time we talk to a customer to say, I had no idea all this amazing innovation that you all are doing, you know, how do we get started? And so we're now having a series of conversations. We do executive briefings and sort of next click down of given where your objectives are, how does this technology enable you to deliver the goals that you have in your business? And so I think the change that the customers are going through in terms of how they perceive Teradata and how they see Teradata on this sort of bleeding edge of innovation, you know, with some of the work that we've done in Gen AI, with some of the large language model work that we're doing and really across the board in terms of a performant multi-cloud data and analytics platform, I think is really been eye-opening and I love hearing feedback from our customers and the pilots that they're doing and how they're getting things into production as well. Excellent. Hillary, thank you so much for coming on theCUBE. This has been great. Thank you, this was awesome, really appreciate it. Thank you. I'm Rebecca Knight for Rob Stretchy. Stay tuned for more of theCUBE's live coverage of Teradata possible.