Oracle Data Deep Dive NYC | Juan Loaiza, Oracle

Clips
More from Oracle Data Deep Dive NYC

Juan Loaiza

EVP, Oracle Database Technologies

Oracle

play_circle_outline Agents require continuous availability, no downtime, 24/7 operation

play_circle_outline Agents as new programming paradigm; specialized agents for data-centric applications

play_circle_outline Oracle positioning database as control plane for enterprise agentic AI

play_circle_outline Autonomous AI vector database: simplified interfaces for data scientists; scalable, secure

play_circle_outline Integrating Quantum-Safe Cryptography into Oracle with Agent-Based Fine-Grained Data Access Controls

play_circle_outline Vectors on Ice: Apache Iceberg support; store and index vectors in data lakes

play_circle_outline Campaign "AI for Data": bring AI to business data with unified, secure foundation

Info
Transcript

Juan Loaiza, Oracle

Juan Loaiza

EVP, Oracle Database Technologies Oracle

In this interview from Oracle Data Deep Dive NYC 2026, Juan Loaiza, executive vice president of Oracle database technologies at Oracle Corp., joins theCUBE's Dave Vellante to discuss how Oracle is positioning the database as the center of gravity for enterprise agentic AI. Loaiza opens with a striking analogy: just as AI can now generate 20,000 lines of code in 20 minutes, the real challenge is no longer building — it's trusting what was built. He explains how Oracle is engineering that trust directly into the database layer, ensuring agents handle data atomi... Read more

explore Keep Exploring

How is Oracle positioning its database as the center of gravity for enterprise agentic AI, and why are mission-critical capabilities and 24/7 availability important? add

What are the recent developments and trends in agentic AI — especially its shift from Q&A to action-taking agents, integration with databases, specialization for data‑centric applications, and improvements in speed, trust, and security? add

What is the "unified agent memory core," and what does it mean for customers? add

What recent updates has Oracle made to its vector capabilities, and how do these changes benefit non-database developers compared with dedicated vector databases? add

How are security challenges from quantum computing and AI being addressed in databases (for example, to prevent encryption breaking and AI-driven data leakage when agents access data directly)? add

How are open table formats like Apache Iceberg and approaches such as "Vectors on Ice" changing data lakes, and how does Oracle support reading, writing, and indexing vector (AI) data in those lakes so that SQL access is seamless across vendors? add

What is your strategy for AI and why are you focusing on bringing AI to business data? add

bolt Powered by CUBE AI

Juan Loaiza, Oracle

search

Dave Vellante

>> My colleague, John Furrier, said in a recent analysis that was covering Oracle's, some of the agentic announcements that Oracle was making in London. He said Oracle's positioning the database as a center of gravity for enterprise agentic AI. So essentially what John was stating is that the future of AI, it's not going to just be determined by the best LLMs or even just the agents, but where and how they interact with data. And that's not even the end of the story because once the agents are up and running, they have to be available. They got to work with data 24/7, 365. Agents don't sleep, so there can't be any downtime. That's not an option. So this is where Oracle's latest mission-critical capabilities and announcements that we saw this week in NYC come in and where AI database fits. Juan Loaiza is here. He's the executive vice president, Oracle AI Database Technologies. He directly reports to Larry Ellison, CUBE alum. You've been here before. Good to see you again. Thanks for coming in.

Juan Loaiza

>> It's great to see you again. It's always fun to talk to you.

Dave Vellante

>> Yeah, we're kicking things off here at our deep dive data event with your support. Really appreciate it. You guys have been at the Javits all week. You had your keynote today. We were talking before about how fast things are changing, how things are different. I wonder if you could kind of summarize what you're seeing out there in the real world.

Juan Loaiza

>> Yeah. I mean, first of all, it's all AI all the time. We all see that. But the world is different than it was even a few months ago. I mean, we had our ChatGPT moment a few years ago where suddenly AI could speak human language and that's amazing and spectacular. And in the last few months, we've had the ChatGPT moment for code where suddenly AI can fluently speak code. Before, you could get code out of it with a bunch of work and you got something. But now, it's just amazing. In 20 minutes, it'll generate 20,000 lines of code for you that actually works. So this is a whole new world we're entering because the limiting factor in the world has been how quickly can you build things? And now, AI can build things that would take many years to build in minutes. And so building things is no longer the challenge. The challenge is, okay, I built this thing, now how can I trust it? Can I deploy it? Can I bet my career or my company on it? That's really the limiting factor.

Dave Vellante

>> Well, the analogy you gave me outside as we were walking in is, Dave, imagine if you could build a building in 20 minutes.

Juan Loaiza

>> There you go.

Dave Vellante

>> Stand up a building. But you asked the question, would you feel safe entering that building? So translate that into technology generally and database specifically.

Juan Loaiza

>> Yeah, exactly. So this is the analogy I use. Suppose I could build 100 story high rise in 20 minutes. Am I going to move my family into there? Well, you better have a lot of trust in the structure of that building and what it's built on. And that's where I'm spending a lot of my personal time on and we're spending a lot of time on is data is the foundation of everything that is knowledge. And so we are building into the Oracle database the ability to trust things that are built on top of it. We can ensure that correct processes, the data is safe. The AI doesn't leak data, it doesn't mismanage the data, it atomically handles the data, it doesn't break business rules with data. So there's basically a whole structure. Think of it like the foundation and the steel beams in the building that ensure that, hey, this thing is safe. And then there's all that stuff around it, which is kind of the app. But without the foundation being safe, everything else doesn't really matter.

Dave Vellante

>> We've known each other for a long time and we've covered Oracle for decades. And it's interesting just observing your company generally and you and your engineers working on database, things change so much in this industry. Rather than going off and sort of pivoting and building sort of separate database, I know you have some. I mean, MySQL HeatWave is obviously purpose built and has its place, but generally you've decided we're going to engineer capabilities into the Oracle database. We don't need to have a separate vector database or a separate graph database. We're going to do that inside of the Oracle database. Now, oftentimes in the technology industry, when companies try to do that, they become less competitive. I think if anything, the opposite is for Oracle, it's become more competitive. I'm curious from an architectural standpoint, how you've been able to pull that off and be able to add these capabilities into a database that I remember the database wars. We were talking about that too from the '80s and '90s. What is it architecturally, culturally, investments in R&D? How have you been able to do that?

Juan Loaiza

>> Okay. So yeah, I mean, what you said is right. We've always believed data is better together. You want to have your data together so you can act on it as once. Atomically with transactions, you want disaster recovery ones, security ones. It's all completely integrated as opposed to, "Hey, I have lots of data that's siloed and broken into pieces, and every time I do something, I have to put it all together. It's slower, less effective, worse answers, all that. Poor security." And one of the things I always say is your security is as good as your worst component and your worst developer. So data's always better together, and that's been our philosophy. And the reason why we're able to do this is, number one, we set this up as a goal. We're not trying to specialize on one thing. We're trying to win the Olympics. We're not in one thing which like just 100 yard dash or something. We need to be general purpose and win the whole Olympics. We need to be better than the specialists at what they're good at. Now, they can sometimes jump ahead because they don't have to deal with everything else. They don't have to integrate it all together, but in a matter of time, we'll catch up and we'll pass them up. And that's what we've done over the years.

Dave Vellante

>> So you're like the decathlete, which maybe is at any snapshot in time, maybe doesn't run the 100-meter dash fastest in the world. But you're saying eventually you become both best of breed across all the different various use cases, maybe with very few exceptions.

Juan Loaiza

>> That's right. Because people think of it as a general purpose database, but we have specific specialized algorithms for each use case and each data type. And then we build that on the foundation that we built over decades, which is the strongest foundation in the world. It runs most of the big businesses in the world, banks, telecoms, airlines, retailers, everything.

Dave Vellante

>> Well, general purpose is important because you can serve more customers and you don't have to slice the salami, as they say, too thinly. Well, you were here almost a year ago now. I think it was the last May in New York City. And then so you're on this world tour now. I definitely paid attention to AI world in Las Vegas last year. It's where you started the whole AI for data revolution with Oracle AI database. You made agentic AI announcements in AI world in London, which expands the capabilities that you guys have. Give us a recap of the recent agentic AI news, if you would.

Juan Loaiza

>> Yeah, sure. So we started with AI a number of years back. And the big challenge then was answering questions, using AI to answer questions, to find data, to answer natural language questions in natural language. So that was the big thing. We have something called vectors that we put into the database to represent the semantic content of data so you can ask questions of your data. So we've done that. The next step in the world is agentic, which is not just answering question, but taking actions and taking actions with trust, actions you can trust. So that's been the big push for the last roughly a year. And it really brings up the whole scale. You can build entire solutions with agents. So it's not just answering questions, it's building entire solutions. And then the underlying fabric of it is really, first of all, making the agents much faster, provide better answers, better security. So we're integrating agents into the database. So agents are like the new programming language, and you don't have specialized programming languages for different purposes. We have specialized agents now for different purposes. Everything gets specialized. Why? Because a specialty programming language is better for the purpose that it's designed for. So we're building an agentic solution for data centric applications. So we're not building it for other purposes. We're not doing it for writing video games or something like that. We're very specialized in data centric applications. And our agents understand the data, understand the database, understand the data security, understand how to manage data, how to process data. And so when you build on top of that, you start out way ahead of the game because all that's built in. The knowledge of what the data is is built into the agents and you can just worry about what you're trying to do with it. So we're specializing the whole thing and we've shipped all that now.

Dave Vellante

>> So speaking to you over the years and just observing Oracle, I mean, I remember Safra would always make this point on our earnings calls is that Oracle has a highly differentiated strategy. You've always talked about the integration of hardware and software. It's Exadata. It's the secret behind Exadata. It gives you certain advantages. So I want to get into the deep dive. This is data deep dive and I want to sort of pick your technology brain. You announced this thing of unified agent memory core. My understanding is it allows for persistent memory. You've got context and state, you've also got long-term memory capabilities, which retain things like user preferences. John Furrier said Oracle strategy is to elevate the database from infrastructure to the control plane for agentic AI. Interesting. I wonder if you agree with that, but I look at it as a converged data model. It's got asset compliance. As they say, you're preserving state. Go give us the double click on unified agent memory core and what does it mean for customers?

Juan Loaiza

>> Yeah. So again, it's a new world and the new world has different challenges and different solutions. So agents, a little bit like humans, like people, they require memory. So there's simple kinds of memory like, this is what you asked me to do last time and here was the answer. You see that when you're using AI normally, but memory gets a lot more sophisticated. It's really kind of knowledge. So the agent, as it runs, it learns things and it has to remember that knowledge. And also, it's not just a private agent anymore. You're working as a team to build things. So that knowledge has to be shared across humans and across agents because now sophisticated systems might use dozens of agents together. And the data, the memory of an agent is not confined to one type of data. It has to know about all the different things. So some things are a document, some things it's trying to remember relational business data. Some things might be text. Other things has to traverse graphs and understand graphs. So that plays into our strength because we're all about unifying the data models, unifying the way you access data, making it extremely fast, making it extremely efficient. And then you really want the memory to be integrated into the actual business data. So if the business data changes, that memory becomes obsolete because the data changed. So you need these things integrated. So it's a whole new world when you're building agents of memory is the real challenge now.

Dave Vellante

>> Okay. So I mean, what you're describing is certainly closer to real time. You're not necessarily going to a data warehouse, which might be hours, days, weeks, or even months old. You're actually making decisions in real time at system speed because agents can move very fast and there's a lot of them. Okay. So that's agent memory core. You talked about vector database. You've got this new autonomous AI vector database. How does it compare with the existing autonomous AI database? How should we think about that?

Juan Loaiza

>> Yeah. So we released our vector capabilities in the Oracle database a couple of years ago.

Dave Vellante

>> Right.

Juan Loaiza

>> And what we've done is we've enhanced that capability, made it much more scalable, available, secure, everything. And now, we're coming out with a specialized version of it, which is really simplified for non-database developers. So people like data scientists or business users that don't want to know about database. They just really want to use the AI, the vectors. So we're providing very simple interface to just the vectors, kind of similar to what a dedicated vector database has. So if you're a data scientist, you don't want anything about database, you don't need to. You have six APIs you use. They're very simple. You can go off and do your vector stuff. And if at some point you hit a wall, you're like, "Well, that was fun with vectors, but now I need to expand to have my AI work against more kinds of data." The underlying database can handle any kind of data. So you can start really simple and you can expand it to the full mission-critical runs everything. So that's really the differentiator because we do the same thing, but better than dedicated vector databases, but we also handle every other thing in the sun. And the goal has really been make it really, really simple for people that don't really want to know anything else about that.

Dave Vellante

>> As an example we were talking about before, can you be best of breed and integrated?

Juan Loaiza

>> Yes.

Dave Vellante

>> And your answer is yes.

Juan Loaiza

>> It takes a little bit of time. It takes a little bit of time.

Dave Vellante

>> Yeah. So you might not be first and all the benchmarks and the leaderboards, et cetera, but eventually you get there. .

Juan Loaiza

>> And at that point, it's not a single stove-pipe solution, you have everything.

Dave Vellante

>> And that's part of the simplicity message. Okay. So you guys, we've said many, many times, mission-critical is you guys are like the alpha of mission criticality. You've built extreme availability, you've got this architecture, MAA, maximum availability architecture. So I want to understand what you've done here. You've announced a platinum level, which comes for free, essentially just an upgrade. I think there's a diamond tier, which is even stock exchange level availability. So what's new in your mission-critical capabilities? Walk us through that for Oracle AI database.

Juan Loaiza

>> Yeah, sure. So there's really kind of two things like you mentioned. So first thing is we have our mainstream database that runs most of the big corporations in the world, governments. So what that's really good at is what I just said, it does everything. It does it with extremely high performance, extremely high security. When you're running a very large database that's super high throughput, multi-node, availability becomes more challenging, right? Because if you have to fail that over, it's like failing over this gigantic thing. But one of the things we've done is we've just improved all our algorithms. So where we had disaster recovery in maybe 50 seconds before, now it might be down to 20 seconds. So we've worked hard at just improving, making everything better. It's like training. You're training for every athletic event. So for example, disaster recovery is in our latest release four or five times faster. Node failures can be up to 10 times faster, much more transparent. We've resolved a lot of issues around rolling patches and that kind of stuff. So that's number one. And four or five times better is a lot. We're not talking 4% better. We're talking four or five times better. And you get that, because our customers get that just by installing our new software. There's nothing they have to do. You don't have to change their application. There's no performance penalty. It runs everywhere. It runs in the cloud. It runs on-premise, on Exadata. In all the leading cloud vendors, you just get it. You install new software, everything is better, faster, better, higher availability, better disaster recovery. And there's another space that we're entering, we call diamond availability, where we're taking a slightly different approach where we're applying, we're copying, we're keeping multiple copies of data consistent through more logical replication and something called charting where we break up the database into multiple databases. So it's a little bit of different approach. It has a lot of benefits also. And one of the benefits is because the changes are being applied kind of a logical level, it's much faster to failover when something happens. So it can go anywhere from zero to three seconds is what we've been measuring. We've been working with a lot of the financial companies to make that literally in the one to two seconds. So really humans won't even notice. It's like, "Hey, you waited one second longer, once every two years when some disaster happened." That's what we're down to. So it's hard to even see that happening. So we've achieved that kind of level of availability for these extreme availability solutions. And we still have our traditional, everything works, everything runs, you don't have to think about anything, you don't have to do anything. It just happens and with extremely low latency.

Dave Vellante

>> I saw that in the press release, Juan, a failover in under three seconds. You just sort of mentioned that, zero to three seconds. But the other piece of it, which really caught my attention was with zero data loss.

Juan Loaiza

>> Yes.

Dave Vellante

>> So that's the hard part, it seems to me. Maybe a lot of times you can fail over, three seconds still pretty impressive. Tell us more about that capability.

Juan Loaiza

>> Yeah. So this is something that you can get by doing what's called synchronous replication. So every time you make a change in one database, you have to make same change at a copy of the database so that if one of them fails, the other one still survives.

Dave Vellante

>> But usually that's slower, right?

Juan Loaiza

>> It is slow. Well, it does slow things down some because it's extra work. You have to go across the network.

Dave Vellante

>> But we're talking three seconds.

Juan Loaiza

>> So there is a trail, which is why we keep both of these technologies around, which is there's a latency trade off. If you're talking sub millisecond latencies, it's hard to do that because of speed of light. So there's a small latency cost to keeping these synchronous copies. And if you want it like thousands of miles distant from the primary database, then there's a larger latency cost. So that's an inherent trade-off in the world. And what we do is we say, "Hey, whatever, you choose," because we can't solve the speed of light, not yet.

Dave Vellante

>> Working on that.

Juan Loaiza

>> Maybe AI can solve it for us. It's a little bit beyond my capabilities today. Maybe with the help of AI maybe next year, we'll come back, we'll talk about it if we do it. But yeah, so customers can choose where they want to live in this spectrum.

Dave Vellante

>> Okay. So you get the platinum tier, that comes for any customer that has the capabilities installed, Exadata, et cetera. And then the diamond tier, that's for a select top of the pyramid, that's a for-pay capability, correct?

Juan Loaiza

>> No.

Dave Vellante

>> No, not necessarily. So, if you have...

Juan Loaiza

>> No. I mean, if you own Oracle database, it's just a configuration. It's a way of configuring .

Dave Vellante

>> So if you have the right configuration, you get that for free as well.

Juan Loaiza

>> Yes. Yes.

Dave Vellante

>> And it's an upgrade.

Juan Loaiza

>> In our latest AI database, there's no new feed features. Any customer that owns Oracle database today owns everything new, all our AI capabilities, our new diamond tier, all that.

Dave Vellante

>> Got it. Are you seeing, because of AI, a renewed interest in mission-critical, is AI become a tailwind? I mean, you used to have mission-critical in your title, so what are you seeing in that dynamic?

Juan Loaiza

>> Yeah. I mean, there's always been an increasing interest in mission-critical, but now what is AI doing? First of all, it's making it much more accessible. You can talk to data in natural language and get answers. You can build applications in seconds to minutes. So the uses of data are going up exponentially because of that. And nobody wants any downtime. Nobody wants any data loss. I mean, there's really no industry really can tolerate that stuff anymore. So yeah, mission-critical is a huge deal, but yeah, I mean, absolutely all this is going to explode the data market. There's no question. There's no question about that.

Dave Vellante

>> I was at RSAC in San Francisco a few weeks ago. Of course, all the talking was talking security. All the talk was about securing agents, some of the concerns that people have about that in addition to prompt injections and data leakage. And so what's your security story? You guys have always been heavily focused on that. Take us through that capability.

Juan Loaiza

>> Yeah. So security has many dimensions and I can talk about a couple. One is quantum computing, security versus quantum computing. We're the first database vendor to introduce quantum safe algorithms in the Oracle database so that even if quantum computers become capable of breaking encryption, they won't be able to break these quantum safe algorithms. So that's one aspect of it. A big aspect is AI brings a whole new set of security concerns. And again, it's not one thing. There's many things that it brings into play. I think the most important part is AI really wants to bypass the traditional application tier. So normally when you interact with whatever, airline system, banking system, medical system, you're going through an interface, right?

Dave Vellante

>> Right.

Juan Loaiza

>> Now, what AI really wants to do is you want to just talk to the system and say, "Hey, tell me about my recent medical test or tell me about how much money I spent last month on groceries." Well, that's not going through the traditional application tier anymore. That's AI going straight to the data and that brings a lot of risks with it, primarily data leakage. So everyone's data is private, whether it's your account data. I don't know what's in your bank account, you don't know what's in mine, I don't know your medical records, you don't know mine, I don't know what flights you're taking. We want to keep that going, right? And so we're having to build a whole layer in the Oracle database, specifically for AI. So AI can act on your behalf. So when you ask, "Hey, tell me what I spent on groceries last month." An AI agent will act on your behalf. You'll ask the question. The AI agent will come to the Oracle database, but when it comes to the Oracle database, we know that it's working on behalf of you, an end user, a consumer, and we will only let that agent see your data. So what you spent, the spend that you did last month, that agent's not going to see my data. And so it cannot leak my data to you or my data out to the internet because you have to log in, you do your normal credentials. The agent acts on behalf of you. It's like a little mini database is created that just has your data in it and nothing else. And so that agent acts against that. It can answer your question without any risk of it leaking anybody else's data. So that's the kind of thing that we're building directly into the database. It comes back to what we said at the beginning. When you build these things in 20 minutes, how are you sure it's not going to leak data? Leaking financial data is a crime. Leaking medical data is a crime. It's not just an optional thing. You go to jail for doing that. So you can't do it. The corporation, it's bad for your career, bad for the company, bad for your freedom. So it has to be guaranteed. And the only place you can really guarantee anymore, it used to be the application tier guaranteed that. That all has to move into the data tier because the AI is operating directly against the data tier.

Dave Vellante

>> So that's a good example of, we always hear in security, you can't just bolt it on after the fact, you've got to design it in. And so that's an excellent example of doing. So we saw with OpenClaw, the inventor of OpenClaw, he was on a long podcast with Lex Fridman, and basically Lex was asking about security. He's like, "Yeah, we'll get there." And as the open source community will solve that, you can't do that in your world. And then of course the quantum safe piece of it, Shor's algorithm is out there. Quantum, so people need to start thinking about that today, not wait until quantum actually becomes mainstream. I wanted to switch topics and ask you about data and data lakes. I saw that you guys rolled out what's called Vectors on Ice, which is a great name. I presume whoever named the Zero Data Loss Recovery Appliance did not name Vector on Ice, so I really liked that. So this expands on your autonomous AI lakehouse capabilities. We've covered that in depth on theCUBE, and SiliconANGLE, I've written about it. What can you tell us about that?

Juan Loaiza

>> Yeah. So a big shift is happening. I mean, we talked about AI a lot and that's a big shift. There's another big shift happening in the data lake world where what used to be proprietary data standards, proprietary data formats are becoming open. So there's a new thing called Apache Iceberg that defines a common table format. So when you store data in a data lake, it'll be stored in this common format that everyone can read and everyone can write. So that's a big thing that we've started supporting. We released support for that starting last year and we're continuing to enhance it. So now, you can load your data into a data lake using this Apache standard iceberg format. Oracle can read and write the data and all of the other database vendors can also, whether it's Databricks, whether it's Amazon, everybody can read and write the same data. So this is something new that's never been possible before. And what you talk about Vectors on Ice is vectors are the new AI data type. So now, you can store vectors in a data lake also in Object store. And what Vectors on Ice are is we can operate against those vectors directly. So we can read the vectors, we can write the vectors. Most importantly, we can index the vectors because the trick with vectors, there's a new kind of indexing taking so you can get at them fast. You don't want to look at every vector in the world. Just like relational data, you want to follow an index so you can get it in sub-second time. So we've made it completely seamless to operate on data that's in the data lake, but also AI data that's in the data lake. So from a point of view of an Oracle customer, and it can be inside the Oracle database, or it can be in this common format outside the Oracle database, and it's completely seamless. There's no difference. The SQL is the same, the way we act on it's the same.

Dave Vellante

>> So what's fascinating to me, Juan, is you could see how somebody might think that Iceberg and OpenTable formats, oh, that's a threat to Oracle because now I can take any database engine and operate on any data. Why would Oracle lean into that? But you choose to lean into that. Obviously, customers, there's momentum there, so you have to pay attention to that. But what's interesting to me is essentially it's made the best engine win. And you probably feel comfortable that you've got a strong, for many, many use cases that you can participate very effectively and competitively.

Juan Loaiza

>> Yeah. So this is our strategy. This is not new. So Oracle started as one of the very first SQL engines. Larry Ellison saw SQL. He said, "Hey, this is the future." And we've helped create that standard over the years and we've helped extend that standard. So we're all about open standards because if we have proprietary things, people don't want to use them. People are like, "Well, it'll lock me into your product." So we've always been a big proponent of open standards and let the best product win, like you said, and we've been doing that forever and we're continuing to do it. And a big area recently was graph database. We were really the leading vendor in standardizing. We put into the ANSI SQL standard the way you access graph data in databases. We've done the same thing with JSON with documents. We were one of the leaders in pushing the syntax, the way you use JSON into SQL into the ISO SQL standard. And similar with Apache standard, we think we're a big believer in standards. That way the entire industry plays together and the entire industry grows.

Dave Vellante

>> So graph is another good example of where you're not saying, "All right, we're going to have a separate graph database." It becomes increasingly important because as we've put forth our vision, which is that organizations want essentially a digital twin of their enterprise, a real-time representation of their business, people, places, things, activities, workflows, processes, transactions. It's all real time. So what does that mean now? Are we entering a world where you're operating on live data where you can sort of talk to that data, get results in real time? How do you see that playing out? What does that mean for you and your customers?

Juan Loaiza

>> Yeah, so we've been about real time data from day one. I mean, we run banks, telecoms, all that is real time. When you make a phone call, that's real time. When you're making a purchase at an e-commerce, that's real time.

Dave Vellante

>> But historically, it's a lot of structured data, right?

Juan Loaiza

>> Yeah. It's been a structured data, but again, we've added support for document data, for graph data, and all that is real time. Everything we do is real time. And, of course, we also support lakes and all that kind of stuff. And the other thing I would add is it's the same product. It's the exact same binary. So the exact same SQL, the exact same data types, the exact same security, the exact same management. It's for all the data types, for all the different workloads. Some other vendors are claiming that they now have OTP and data lake together, but it's two separate products that weren't designed together. They use different SQL engines, they use different data types, they have different security, different management, but they're claiming they have this unified product. It's not unified. It's a completely different thing. So we're really the vendor that supports everything from real time OTP to giant data lakes, from relational to graph, to JSON, to anything you want all in one unified product. And it all works completely seamlessly together because it's one product.

Dave Vellante

>> Impressive. My last question, you've got a new campaign around Oracle AI database. You guys start with engineering and then obviously you market it. Sometimes you market ahead of time just to get the market ready. But what can you tell us about the new campaign?

Juan Loaiza

>> Yeah. So we're promoting the concept of AI for data that you should bring AI to your data. We're bringing vectors, we're bringing all that. Why? Because you really need to do AI with your business data. The business data is the ground truth of the enterprise. It says what you're selling, who your employees are, what your products are, what your manufacturing is happening. And you have to apply AI together with that data to get business results. So that's the number one thing we're saying. Bring AI to the data. We've integrated the whole thing together. We've made it seamless. You can do a simple SQL queries. We have specialized agents. We make it easy to create applications. We make it safe to use the data. We make it secure. That is the biggest part of it. And in addition, we're also continue to talk about, because we continue to improve all our mission-critical technologies, security technologies, integrating all the data, everything we've done over the years. But yeah, AI is a huge focus. AI is clearly the big new thing. And we want to be the foundation for building AI and we think we have really huge advantages in the market for that.

Dave Vellante

>> Well, it's awesome to have you here. We have a lot of Oracle customers in our community when we saw you guys were doing this event at Javits. We were actually originally going to come to the Javits, and we said, "Hey, you think you could bring your team down here?" So I really appreciate you spending some time with us and making your team available for this program. So thank you.

Juan Loaiza

>> I always have a lot of fun talking to you. .

Dave Vellante

>> Thank you. I appreciate that. All right. You're watching Oracle Deep Dive in New York City. I'm Dave Vellante. You're watching the NYSE Wired and theCUBE's continuous coverage. Keep it right there. We'll be back with more great content right after this.