 Welcome back everyone to theCUBE's coverage here on the location in Las Vegas for AWS re-invent. I'm John Furrier, host Dave Vellante's here. So the analyst session is going to come on for a wrap up here today. But a whole team coverage here in Palo Alto and in Las Vegas bringing super qualified special edition battle for AI supremacy. This is how the next generation infrastructure supporting next generation AI apps are here. Generative AI is hot. That's the topic. We've got two great guests coming on here. Haseeb Bhudhani who's the CEO of Rafi Systems, CUBE alumni. Great to see you back again. Michael Cuell who's the AABG Global Container Lead for Accenture. AABG stands for Accenture's AWS Business Group. Accenture's got great titles. Always loved the title of Accenture. Welcome both of you. Thank you, we're glad to be here. Great to see you Steve. So talk about the Kubernetes situation because this has been one of the fast evolving situations. We had many conversations at Supercomputing in Denver and here in India-Various about Kubernetes being involved in standing up clusters. So big customer change is becoming standard as they say at CUBECon. Kind of boring which means it's working but the value is shifting on the front end for more stuff. So you're seeing customers standing up clusters. The big customers whether it's GPUs and what's around the chips. That was a big talk of Adam Silevsky's keynote. It's not just about the chips or the models. It's what's around them, choice. These are like, this is like a very key inflection point right now for the industry. Yeah, I would say that at this point maybe the A and AWS is also AI, right? Because everything is AI these days. Look, you and I talked at CUBECon about Kubernetes becoming boring, right? Which means become a standard. And I think that that idea was sort of more further clarified for us here at Reinvent because all the customers we're talking to they're validating a perspective we've had for the last six months or so. Which is, let's not talk about Kubernetes anymore, let's talk about outcomes. You Mr. Platform Guy, what are you trying to do with Kubernetes? What is your end customer need? And as I'll give an example, we have a customer who came to us three months ago and said this bedrock thing, everyone should try it. I want to make it available as a service to my internal development team. How do I do it? So, okay, that's an interesting outcome. It needs Kubernetes by the way. But they never said the word Kubernetes, right? And I love that about this, right? Because now we're at a point where people are going beyond their technology and thinking about what does my customer actually need to be successful? And what guardrails are going to need to put in place so that they don't go crazy, obviously. Spend a lot of money. That's where we are. What's great about this is the natural level is when we first met, I don't know, if you're at the first time, we're riffing on like where Kubernetes is going to go, is it the middle layer orchestration, TCP IP for the modern era, cloud native kind of metaphor. Michael, this is now showing the use cases when you start to get into practice. Okay, we're setting up real engineering or oriented environments, data engineering, platform engineering. These are environments that are being engineered now for these outcomes. And the stakes are high. Everyone knows what the stakes are. It's general AI applications, a feeding frenzy for developers to be feeding on either traditional software development tactics and or the new foundation models, which will open up more agility. And you know, Sileski said on stage, adaptability is the key to winning this. Whether you're a customer or a vendor, the use cases drive everything. Right, exactly. That's what we see from our customers. What are they trying to do? What are they trying to solve? As they're trusted advisors, we bring solutions to the table based on their needs, right? So whether that's gen AI or app modernization, or like we're seeing a lot of this now, is mainframe modernization. Getting away from them, COBOL programmers that are retiring and sun setting those applications and bringing them into Java and putting them into Kubernetes. You know, that's a really nice time to be working. You can write a chatbot for COBOL. Ingest the language, train it, and then it furs and thing. COBOL is like, you don't have to hire anyone. But we saw great things in the keynote. Like I saw the thousand applications converted in two days, incredible numbers. And then the telegraph of .NET to Linux migrations. The AI will start doing some of these things with Q and whether it's Code Whisperer and Q together. This is going to open up opportunities. And that's why I want to get into the relationship between you guys, Centra and Rafi system, because your collaboration is spawning new observations with customers. What are some of those things that you're seeing? Because as Kubernetes falls in the background as a standard, the enablement that is, it's a disruptive enabler. It's enabling more things to move faster. And on top of it already hyped up, obvious JNAI market. What are the use cases? What problems that customers having? What are some of the things you guys are solving together and what are you observing? You want to take that? Sure. Look, both of us were just in a meeting right before this where I asked the customer, hey, what are you looking for? And the good news is customers like a year ago, or three years ago when we started talking about Kubernetes, they would say things like I'm looking for e-gas or I'm looking for a policy engine, right? This lady said, I'm looking to deliver cluster service, namespace service, landing zone as a service to my internal customers. So much clarity, right? This is great for us as a vendor because it just makes our life easier. But I will tell you that the reason why we are here, us as Rafi, and we've been able to deliver on those packages, if you will, is because of Accenture, right? Have you spent a lot of time together over the last year? So we've been working on this partnership for a year now. Accenture's a big company, it takes time. But we are at a point now where, you know, they are our chosen GSI partner. We've trained a lot of their people on our platform as one example. But the value here, hopefully they see value, but the value I see is, man, they understand how to have the strategic conversation with their customers, right? Really help the customer, don't talk about the technologies, let's talk about what are you trying to solve for? The word outcomes, every view's already in this conversation, right? And when we start talking about outcomes, everybody wins, right? Obviously we as a vendor, we want to make money, we want to sell our product, but the outcomes make the customer successful. And Accenture seems to have a very unique way of making that sort of come out, if you will, from the customer side. Tell us about some of those conversations because when I hear outcomes, okay, I'm not rolling my eyes, believe me, and when I say this, I'm just compliment. There's still stuff under the covers, right? And one of the themes this year has been amplified. Here and in other events, it's end to end. The workflow, here workflows, because AI affects everything. So, okay, work backwards on the outcome. There's a lot of stuff under the covers that you guys are involved with the customer on to translate the puzzle pieces of the architecture, because there's a lot of architecture change going on, right? So take us through some of the outcomes and what systems need to be in place for those outcomes, that's new. Wow, so security, gen AI fitting into security, like there's a huge concern with cyber malware. Like, they don't want to see their data corrupted and being held for ransom, so that's not an outcome, but that's a preventative measure that we have to take to get to their desired outcome. We have to see that they're delivering these applications in a secure manner. They're not getting any threats from cyber work, because many of our customers, Fortune 100, they've got big targets on their back. Their household names, the cyber threats out there, it's real, ransomware is very real. How do we protect them from that? Let's bring the AI components in, let's deliver that through partnership with Rafay and their environment manager, and let's set this thing up so that it's bulletproof, right? So that's one of the ways we get to that outcome. And then, obviously, the security side, I mean, the bad guys have no low bar of testing anything, they try anything. High bar for app sec review and getting things tested in security, in complex AI workloads, it's a high bar. You guys have a high bar to deliver. Bad guys got you into AI technology too. Yeah, and we're the trusted advisors, our word means a lot to our customers. They listen to us, they come to us and say, hey, help us solve these problems. How do we do it? How do we get there? So we have to deliver excellence, that's our goal. What are some of the things you guys learn in your partnership, specifically around use cases? What ransomware is obviously one, security's one, what other things on the platform engineering side are you seeing? I'll give you a boring example. So let's say a developer needs to do some rack testing on some level. How much does it cost? Because I'm the manager, I'm not going to approve this environment till I know how much it costs. So what if you could build a template, and we build this by the way, the product does that by the way, so you say I want to launch this specific use case with Bedrock. I'm using Bedrock because we're, of course, we're behind right now. And the platform should be able to say, okay, Mr. Manager, please approve this, and it's going to cost $2,000 a month and the manager can make a decision. Am I going to approve it or not? Boring, but everybody cares because there's not one developer, there's 1,000 of these people. If every developer spun up a GH100 or 200, man, that's 20 grand a pop per month. That's a lot of money. These are the kinds of things that we've learned over the last year that are really, really important to customers, that for them to bring this all together in one solution is very, very hard. And yeah, look, I mean, we've learned this in partnership with our friends. And that's not a boring example. That's a legit, that's a real life example. It's money, it's like technology. Well, I mean, what we're hearing here at this event, I mean, cost and price performance is now on the table. Absolutely. The infrastructure is now being tested because to match the security latency, both latency in terms of packets and time to value on inference, for instance, is huge. The new latency is not, it's your next answer. It's the next thing. It's not just packets. It's the AI latency. So we're seeing people talk about how fast can you get to proof points that I'm going to go to the next level, not, okay, here's some nice web copy from my marketing page. I mean, that's a good example, but it's trivial compared to some of the harder AI opportunities where it's like, I want to start getting teams solving a problem. We had an example from Eric, from the security group at Amazon, on one hackathon, they took the knowledge base, crawled it, and they had a workable product. It wasn't great, but it was good enough for the engineers saying, I like this. I see value. And now they're all got FOMO in on it. And now they're working together. So we're in a kind of a collaborative, like you can see value quicker as a developer or as a solution provider. And it doesn't have to be perfect, but you're like, okay, I see enough to go the next step, versus maybe a week, I see something to get people to meeting, go build it, go proof it. Now it's instant value. Yep, right, absolutely agree. Look, our initial co-pilot, everybody's got a co-pilot, we have one too. It took us, I think, a night to build, though at least the first implementation to say, this is interesting, we should invest more, and we are investing more. And that's something that we could not do before, of course, right? Now we can do this with the app. And the value of the data, right? So one of the things that came out of super computing and here again, it's a little bit more super computing, but this is cloud here at Amazon, is using Kubernetes for purpose-built infrastructure. Maybe a GPU cluster, especially scope to match to, say, a service for developers. This is kind of what we're seeing as the trend. You mentioned model service, the other example. How much of that's going on in your accounts? Interesting you say that. One of the things we love to talk about is accelerators. Get things done faster. Reusable code, reusable modules, things that we can stand up environments and take these LLMs and bring them into play so you can run your routines and your bedrock foundation and look at this data and ingest it and speed everything up. That's just the way that we approach it. We love reusing accelerators. Your environment manager is going to be a big part of that where we'll bring all of our tarform modules or cloud formation modules, or you cast blueprints into the picture and be able to deploy the speed, right? That's our goal here. Great. So what else is new, guys? What do you think about the keynotes? What about the event? I mean, it's packed, feels like 2019 when I think the high water mark was 60,000. I think... How many people is it this year? 65K? Something that maybe war seems like it's... I think they were expecting 50,000 but I think they pushed over 60 easily. It's definitely the most crowded I've seen. It's a zoo. It's definitely a zoo. I love the balls in the way up. I think that's to prevent road rage. Because I'm doing a full core body workout and just dodging the traffic in there. If you want to walk faster to that crowd, it's like a workout. It's like a... We just have friends. Yeah, there's donuts on the floor. If you eat those, you got to eat those anyway. Which I have. I have it a couple of days. Well, what are you guys seeing? What's the big walk away? What's the most important story happening here? It's AI. It's AI. I mean, AI is for sure. I think what I was thinking about earlier today and John, I'm sure you have a strong opinion on this. When we're here next year, God willing, how important will AI be at that time? Will it still be the same sort of vigor and tenor? Or will it become part and parcel of everything we do? Right? And I don't know right now. I mean, I know what our business is going to have to do with it, but what do you think? Well, I had an hour with Adam Sileski on November 17th as a preview for the article I wrote on Sunday night that's for the curtain raiser exclusive. And that was this, when the interview was over, an hour after it was over, is when Sam Altman got fired. And the drama then in 2000 Seattle for that next day. And I was kind of watching the whole play-by-play train wreck and they got to be hired back. But what that changed was to me, was the culture of, oh my God, that's going to put Microsoft sideways for a while. And it did. Advantage to Amazon, he kind of brought up his keynote. But to me, I think what I see clearly is that there's going to be a developer feeding frenzy on the LLM layer. So I think the three layers is legit. I think he's nailed it as a GNAI stack. I think that middle layer is going to be a shim layer between these kind of lightweight capabilities that's going to enable the ecosystem to add value. And I think they're clearly not going down the Microsoft path of we're going to have our own apps and instantly put co-pilot in there and charge 30 bucks and show success. I think that's good for Microsoft to do because they can do it. That's all that they only move they have. And then hope that no one uses GNAI to recreate word in Excel because that's what's going to happen next, right? So that's one. In terms of next year, I think this is the web in my generation. That's what I relate it to. The VM where people relate it to the virtualization wave if they're in that generation. But to me, it's going to be all about what's improved. GNAI is here. It's like web pages loading in 2019-95, slow dial-up modems, it then slow gets fixing, it's faster, modems gets faster, everything's getting better than the next processors come out from Intel. So I think we're going to see a continuation of that same kind of more price performance, better energy. The constraints are no longer the motherboard, like a PC. It's the constraints are power, creativity. And I think you're going to see a tsunami of new applications. And I think it's going to be next year, who's got better stuff? Who's got better speed? And what is the sense you're doing for more outcomes? What is, what app, what is perplexity doing for their next app if they're on stage? Or how do I eliminate data pipelines completely from the development process? I mean, these are the things I think, things that you're working on. It's like, okay, your job should get easier where you're invisible. Yeah, you know, that would be like, okay, then you're successful. Yeah, I would love to love the forecast. Find things like perplexity. I heard about perplexity from my 11-year-old. Because he uses it. Amazing. I love that app. I love that app. And the human skill set is changing. The AI is forcing that. They're coming up with new skill sets, more investment in our talent. And how do we continue to deliver values to our clients? That's morphing as Jenna takes more of a focus. It's an interesting time for sure. I mean, no one's really talking about what we've been seeing on the queue all the time, but data also lives in our head. As human beings. And I think one thing that this does is scales intellect. So intellect is also what we have as humans. And I think that piece of the puzzle is going to be something that's going to be a creative era of creativity we've never seen before. I'm going to talk about creative. I'm talking about all corners of technology. To the configurator, to the operator, to the developer, to the basic worker. I mean, that's why Q is so cool. It's like their copilot. Okay, they finally got it. So I think we're going to start to see every layer of society just unleashing something. Hopefully it's good, not bad, because the hackers are getting it too. So yeah, we don't want war games. That's for sure. Guys, thanks for coming on theCUBE and sharing your opinion. Great to see you again. Congratulations on your partnership with Ascension. Thanks for having us. Good stuff. Thanks for having us. Cube coverage here. Back to the studio. We'll be back with more coverage here in Las Vegas after this short break.