 Good evening, everyone, and welcome back to Sparkly, Sin City, Las Vegas, Nevada, where we're here with theCUBE, covering AWS re-invent for the 10th year in a row. John Furrier has been here for all 10. John, we are in our last session of day one. How does it compare? I just graduated high school 10 years ago, exciting to be here. Been a long time, gotten a lot older. Oh my God. God, your brain is complex. You've been a lot in there so fast. You guys were in eight in high school, you know how it is, but all good. What's going on in this next segment? Wrapping up day one, which is like the kickoff of Monday, it's great here, I mean Tuesdays, coming tomorrow, big days. The announcements are all around the kind of next gen and you start to see partnering and integration is a huge part of this next wave because APIs is the cloud, next gen cloud's going to be deep engineering integration and you're going to start to see business relationships and business transformation scale horizontally, not only across applications, but companies. This has been going on for a while, but we're covering it. This next segment is going to be one of those things that we're going to look at as something that's going to happen more and more. Yeah, I think so. That's what we've been talking about all day. Without further ado, I would like to welcome our very exciting guest for this final segment. Dress from Single Store. Thank you for being here. Thank you. And we also have him on from IBM, Data and AI. You all are partners, been partners for about a year. I'm going to go out on a limb only because they're legacy and suspect that a few people, a few more people might know what IBM does versus what a Single Store does. So why don't you just give us a little bit of background so everybody knows what's going on? Yeah, so Single Store is a relational database. It's a foundational relational systems. But the thing that we do the best is what we call as real-time analytics. So we have these systems that are legacy which do operations or analytics. And if you wanted to bring them together like most of the applications want to, it's really a big hassle. You have to build an ETL pipeline. You have to duplicate the data. It's really faulty systems all over the place. And you won't get the insights really quickly. Single Store is trying to solve that problem elegantly by having an architecture that brings both operational and analytics in one place. Probably. You guys had a big funding now, expanding MemSQL, Single Store, databases. 146 billion. Again, databases, we've been saying this in the queue for 12 years, been great. And recently, not one database will rule the world. We know that, everyone knows that. Databases, data, code, cloud scale. This is the convergence now of all that coming together where data, this reinvent is the theme. Everyone will be talking about end-to-end data, new kinds of specialized services, faster performance, new kinds of application development. This is the big part of why you guys are working together. Explain the relationship, how you guys are partnering and engineering together. Yeah, absolutely. I think so, IBM, I think we are mainly into a hybrid cloud and AI. And one of the things we are looking at is expanding our ecosystem, because we have gaps. And as opposed to building everything organically, we want to partner with the likes of Single Store, which have unique capabilities that complement what we have, because at the end of the day, customers are looking for an end-to-end solution that solves their business problems. And they are very good at real-time data analytics and a head-stab, right? Because we have transaction databases, analytical databases, data lakes, but head-stab is a gap that we currently have. And by partnering with them, we can essentially address the needs of our customers. And also what we plan to do is try to integrate our products and solutions with that so that we can deliver a solution to our customers. This is why I was saying earlier, I think this is a tell sign of what's coming from a lot of use cases where people are partnering. Right now you've got the cloud, there's a bunch of building blocks. If you put it together yourself, you can build a very durable system, very stable. If you want out-of-the-box solution, you can get that pre-built, but you really can't optimize it, break it or replace it at high level. Engineering systems together is a little bit different, not just buying something out of the box. You guys are working together. This is kind of an end-to-end dynamic that we're going to hear a lot more about at re-invent from the CEO of AWS. You guys are doing it across companies, not just with AWS. Can you guys share this new engineering, business model, use case? Do you agree with what I'm saying? Do you think that's crazy? Do you think John's crazy? I mean, we're open all discourse here. You've got to out-of-the-box engineer yourself, but then now when people do joint engineering projects, they're different. Yeah, yeah. You know, I think our partnership is a testament to what you just said, right? When you think about how to achieve real-time insights, the data comes into the system and the customers and the new applications warn insights as soon as the data comes into the system. So what we have done is basically build an architecture that enables that. We have our own storage and inquiry engine indexing, et cetera, and so we've innovated in our indexing, in our database engine. But we want to go further than that. We want to be able to exploit the innovation that's happening at IBM. A very good example is, for instance, we have a native connector with Cognos, their BI dashboards, to recent data very natively. So we built a hyper-efficient system that moves the data very efficiently. A very other good example is embedded AI. So IBM, of course, has built AI chip and they have basically advanced quite a bit into the embedded AI, custom AI. So what we have done is, as a true marriage between the engineering teams here, we make sure that the data in single store can natively exploit that kind of goodness. So we have taken their libraries, so if you have data in single store, like let's imagine, if you have Twitter data, if you want to do sentiment analysis, you don't have to move the data out, drain the model outside, et cetera. We just have the pre-built, embedded AI libraries already. So it's a pure engineering manage there. That kind of opens up a lot more insights than just simple analytics. And cost, by the way, too, moving data around. Another big thing. Yeah, and latency and speed is everything about single store and it couldn't have happened without this kind of a partnership. So you've been at IBM for almost two decades. Don't look at it, but at nearly 17 years in, how has, and maybe it hasn't, so feel free to educate us, how has IBM's approach to AI and ML evolved as well as looking to involve partnerships in the ecosystem as a collaborative, raise the water level together, force? Yeah, absolutely. So I think when we initially started AI, I think we had a few recollects. Watson was the fourth friend of AI. We started the whole journey. I think our focus was more on end solutions, both horizontal and vertical. Watson Health, which is more vertically focused. We're also looking at Watson Assistant, Watson Discovery, which were more horizontally focused. I think that whole strategy has evolved over a period of time. Now we are trying to be more open. For example, this whole embedded AI that Sirish was talking about, it's essentially making the guts of our AI libraries, making them available for partners and ISVs to build their own applications and solutions. We've been using it historically within our own products for the past few years, but now we're making it available so that- How big of a shift is that? Do you think we're seeing a more open and collaborative ecosystem in the space in general? Absolutely, because I mean, if you think about it, in my opinion, everybody is moving towards AI and that's the future, and you have two options. Either you build it on your own, which is going to require a significant amount of time, effort, investment, research, or you partner with the likes of IBM, which has been doing it for a while, right? And it has the ability to scale to the requirements of all the enterprises and partners. So you have that option, and some companies are picking to do it on their own, but I believe that there's a huge amount of opportunity where people are looking to partner and source what's already available as opposed to investing from the scratch. Classic buy versus build analysis for them to figure out to get into the game. And why reinvent the wheel when we're all trying to do things at not just scale, but orders of magnitude faster and more efficiently than we were before, it makes sense to share, but it does feel like a bit of a shift, almost a paradigm shift in the culture of competition versus how we're going to creatively solve these problems. There's room for a lot of players here, I think, and yeah, it's, I don't know, it's really... I wanted to ask if you don't mind me jumping in on that. It's okay, I get that, people buy build, I'm going to use existing or build my own. The decision point on that is to point about the path of getting in the path of AI is, do I have the core competency? Skills gap's a big issue. So okay, theCUBE, if you had AI, we'd take it because we don't have any AI engineers around yet to build on all the linguistic data we have, so we might use your AI, but I might say to Savannah, we want to have a core competency. How do companies get that core competency going while using and partnering with AI? What do you guys see as a way for them to get going? Because I think some people will probably want to have core competency. Yeah, so I think again, I think I want to distinguish between a solution which requires core competency, you need expertise on the use case and you need expertise on your industry vertical and your customers versus the foundational components of AI which are agnostic to the core competency, right? Because you take the foundational piece and then you further train it and define it for your specific use case. So we're not saying that we are experts in all the industry verticals. What we are good at is foundational components, which is what we want to provide. Got it, yeah. That's the hard, deep, heavy lift. Yeah, and I can give a color to that question from our perspective, right? If you think about what is our core competency, it's about databases, right? But there's a symbiotic relationship between data and AI. You know, they sort of like really move each other, right? You need data. They kind of can't have one without the other, right? And so the question is, how do we make sure that we expand that relationship where our customers can operationalize their AI applications closer to the data, not move the data somewhere else and do the modeling and then training somewhere else and dealing with multiple systems, et cetera. And this is where this kind of a cross-engineering relationship helps. Awesome, awesome, great. And then I think companies are going to want to have that baseline foundation and then start hiring in, learning. It's like driving a car, you get the keys when you're ready to go. Yeah. Yeah, I think I'll give you a simple example, right? I want that turnkey lifestyle, we all do, yeah. Yeah, let me just give you a quick analogy, right? For example, you can basically make the engines and the car on your own or you can source the engine and you can make the car. So it's basically an option that you can decide. The same thing with airplanes as well, right? Whether you want to make the whole thing or whether you want to source from someone who is already good at doing that piece, right? So that's- Are you even creating a new alloy for that matter? I mean, you can take it all the way down in that analogy. Is there a structural change in how companies are laying out their architecture in this modern era as we start to see this next-let-gen cloud emerge? Teams, security teams, becoming much more focused, data teams, ITs building into the DevOps, into the developer pipeline, seeing that trend. What do you guys see in the modern data stack kind of evolution? Is there a data solutions architect coming? Do they exist yet? Is that what we're going to see? Is that data as the code with automation? How do you guys see this landscape of the evolving persona? I mean, if you look at the modern data stack as it is defined today, it is too detailed, it's too verbose, and there are way too many layers, right? There are at least five different layers. You've got to have like a storage, you replicate to do real-time insights, and then there's a query layer of visualization and then AI, right? So you have too many ETL pipelines in between, too many services, too many choke points, too many failures, right? ETL, that's a dirty three-letter word. Say no to ETL. Adam Slessy, that's his quote, not mine. We'll hear that tomorrow. Yeah, I mean, there are different names to it. They don't call it ETL, they call it replication or not. But the point is, the data is getting more hassle. More hassle. The data is ultimately getting replicated in the modern data stack, right? And that's kind of one of our thesis at Single Store, which is that you'd have to converge, not hyperspecialize, and convergence is possible in certain areas, right? When you think about operational analytics as two different aspects of the data pipeline, it is possible to bring them together and we've done it. We have a lot of proof points to it. Our customer stories speak to it. And that is one area of convergence. We need to see more of it. The relationship with IBM is sort of another step of convergence, wherein the final phase is the operation analytics is coming together, and can we take analytics visualization with reports and dashboards and AI together? This is where Cogloads and Embedded AI comes into together, right? So we believe in Single Store, which is really convergence. One single path. A shocking tie back there. So obviously, one of the things we love to joke about on theCUBE, because we like the goof on the old enterprise is they solve complexity by adding more complexity. That's old thinking. The new thinking is put it under the covers, extract away the complexities, and make it easier. So how do you guys see that? Because this end-to-end story is not getting less complicated. It's actually, I believe, increasing in complexity. However, there's opportunities to put it under the covers or put it under the hood. What do you guys think about how this new complexity gets managed or in this new data world we're going to be coming in? Yeah, so I think you're absolutely right. The world is becoming more complex. Technology is becoming more complex. And I think there is a real need. And it's not just coming from us, it's also coming from the customers to simplify things. So our approach around Embedded AI is exactly that, because we are essentially providing libraries. Just like you have Python libraries, Java libraries, now you'll have AI libraries that you can go infuse and embed deeply within applications and solutions, so it becomes integrated and simplistic. For the customer point of view, from a user point of view, it's very simple to consume, right? So that's what we're doing. And I think Single Store is doing that with data, simplifying data, and we are trying to do that with the rest of the portfolio, specifically AI. It's no wonder there's a lot of synergy between the two companies. John, do you think they're ready for the Instagram challenge? Yes, they're ready. I think they're ready. So we're doing a bit of a challenge. Little 30 second off the cuff. What's the most important takeaway? This could be your, think of it as your thought leadership sound bite from AWS 2022. On an Instagram reel, I'm scrolling. That's the Instagram reel. It's your moment to stand out. Yeah, exactly. All right. Suresh, you look like you're ready to rock. Let's go for it. You've got that smile. I'm gonna let you run. Oh goodness. You know, there's this quote from Astrophysics. Space moves matter. A matter tells space how to curve. They have that kind of a relationship. I see the same between AI and data, right? They need to move together. And so AI is possible only with right data and data is meaningless without good insights through AI. They really have that kind of relationship. And you would see a lot more of that happening in the future. The future of data and AI are combined and that's going to have an accelerate a lot faster. Suresh, well done. Thank you. I am very impressed. Hey man, tough acts to follow. You ready for it though? Let's go. Absolutely, yeah. So just to add to what Suresh said, right? I think there's a quote from Rob Thomas, one of our leaders at IBM. There's no AI without IA. Essentially, there's no AI without information architecture which essentially is data. But I want to add one more thing. There's a lot of buzz around AI. I mean, we're talking about simplicity here. AI in my opinion is three things and three things only. Either you use AI to predict future for forecasting, use AI to automate things. It could be simple mundane tasks, it could be complex tasks depending on how exactly you want to use it. And third is to optimize. So predict, automate, optimize. Anything else is buzz. Okay. Brilliantly said. Honestly, I think you both probably hit the 30 second time mark that we gave you there and the enthusiasm left your hunger on that. You were born ready for that kind of pitch. I think they both nailed it for this. They nailed it. Nailed it. Well done. I think that about sums it up for us. One last closing note and opportunity for you. You have a V8.0 product coming out soon December 13th if I'm not mistaken. You want to give us a quick 15 second preview of that. Super excited about this. This is one of our major releases. So we are evolving the system on multiple dimensions on enterprise and governance and programmability. So there are certain features that some of our customers are aware of. We have made huge performance gains in our JSON access. We made it easy for people to consume wasm on on-prem and hybrid architectures. There are multiple other things that we're going to put out on our site. So it's coming out on December 13th. It's a major next phase of our system. And real quick wasm is the web assembly momentum. Correct. The new. Yeah, we are pioneers in that. We bet it wasm inside the engine. So you could run complex modules that are written in, could be CE, could be Rust, could be Python. Instead of writing the SQL, anti-SQL as a store procedure, you can now run those modules inside. I wanted to get that out there because at KubeCon we covered that Savannah. It's a hot topic. Like a blanket. We covered it like a blanket. Wow. On that glowing note. Suresh, thank you so much for being here with us on the show. We hope to have both single store and IBM back on. Plenty more times in the future. Thank all of you for tuning in to our coverage here from Las Vegas in Nevada at AWS re-invent 2022 with John Furrier. My name is Savannah Peterson. You're watching theCUBE, the leader in high tech coverage. We'll see you tomorrow.