Exclusive Coverage: theCUBE at IBM Think | Varun Bijlani & Hillery Hunter | IBM

theCUBE.net

Clips
News
More from Exclusive Coverage: theCUBE at IBM Think

Varun Bijlani

Global Managing Partner - Hybrid Cloud Services

IBM

Hillery Hunter

CTO, IBM Infrastructure

IBM

Three insights you might have missed from theCUBE’s coverage of IBM Think

IBM Corp. was an early adopter of AI with its “Jeopardy!”-winning Watson computer. Now, Big Blue looks to make a comeback by positioning itself as the “glue layer” for enterprises trying to integrate AI across platforms.At IBM Think 2024, the company announced a series of products based on its AI platform, watsonx, and showcased its dedication to developing a network of partnerships. Its acquisition of Red Hat Inc., a leader in open-source solutions, has continued to pay off. IBM is expanding on that ethos of collaboration by releasing Granite models into open source

Hybrid AI: The convergence of hybrid cloud and artificial intelligence

As artificial intelligence expands to cover an increasing number of industries, an emerging topic is hybrid AI — which combines the hybrid cloud model with AI’s data processing capabilities.What constitutes hybrid AI practically and how (if at all) is it poised to add concrete value to the existing AI transformation wave?TheCUBE explores IBM’s hybrid AI strategy with Varun Bijlani and Hillery Hunter at IBM Think 2024.“The first fundamental thought out here is that there is a very significant symbiotic relationship between hybrid cloud and AI,” said Varun Bijlani (pictured, left), global managing partner of hybrid cloud services at IBM Corp

play_circle_outline The Intersection of Hybrid Cloud and AI: The Crucial Role of Unified Platforms and Overcoming Challenges

play_circle_outline Unleashing Enterprise Data and Accelerating Hybrid Cloud Execution: IBM's Advanced Hybrid AI Capabilities

play_circle_outline Consistency of developer and operational experience in hybrid cloud and AI

Info

Varun Bijlani & Hillery Hunter | IBM

Varun Bijlani

Global Managing Partner - Hybrid Cloud Services IBM

Hillery Hunter

CTO, IBM Infrastructure IBM

search

>> Hi, everybody. Welcome to the special CUBE presentation. We're here at the BECC in Boston where theCUBE started in 2010, the first CUBE ever at EMC World in 2010. Really excited to have this special presentation here. Inside the IBM studio, it's fantastic. Varun Bijlani is here. He's the Global Managing Partner for Hybrid Cloud Services at IBM. And Hilary Hunter, who's the CTO and GM Innovation, love that title, for IBM Infrastructure. Folks, welcome back to theCUBE-ish.

Varun Bijlani

>> Thanks so much.

Varun Bijlani

>> Thank you so much.

>> Thank you.

>> Great to see you guys. So last week we put out a post on our Breaking Analysis, talking about the sort of intersection between hybrid cloud and AI, what we called hybrid AI. We clearly see these two things coming together in a different pace. And so, I wonder if you could share your thoughts on what you're seeing from customers going around this show at IBM think. What are you hearing? What are you seeing? What do you expect for the intersection of those two concepts?

>> Yeah, I definitely want to hear a little bit more about your data. But I think just to share some of ours, if you look back like six or seven years ago, hybrid cloud wasn't so much a thing. Cloud was sort of a place where you went to do your computing, where you went to use your AI. And I think the reality from our polling is, today, if you ask CIOs how much their cloud estate represents hybrid cloud, about 70% of them will say that they actually have a hybrid cloud infrastructure. So hybrid cloud is a thing, it's a thing everyone is doing. But it often when you dig into the details, is actually a bunch of disparate and heterogeneous cloud environments, and not a unified place that you really have a platform to create new things.

>> So, thank you for that. Varun, I think in some ways, IBM maybe has an unfair bias, and here's what I mean by that. What our data shows is that, and I think this has become sort of a bromide, bringing AI to the data and a lot of data on-prem. I think a lot tons of enterprise data, if you take out video for consumers is on-prem. What I mean by the unfair bias is in order for hybrid AI to happen, you've got to have the tooling, you've got to have the LLM optionality. You've got to have the ecosystem. Actually, IBM has those things. That's why I say it's sort of an unfair bias. A lot of the sort of traditional on-prem vendors maybe don't have that. I wonder if you could comment on how you're bringing that all together.

Varun Bijlani

>> I think the first fundamental thought out here is that there is a very significant symbiotic relationship between hybrid cloud and AI. And today, AI is supremely accelerating the execution of hybrid cloud. And at the same time, you can't really go from pilot to production to scale without a robust architecture. So, I think that's the first reason why this conversation right now is so relevant. And then in terms of our own independence, our thought process around this is about making sure that we are allowing you to accelerate the value of your own data. Up until now, all the conversations have been about, I've got models with large amount of public data, but the amount of enterprise data in there is next to negligible. Your differentiation is going to come when you can take your data and use that with those right models. And I think that's where we've decided to make our conversation much more open. And you heard today, Granite is now open-sourced. And then you start saying, "Where do we have the unfair advantage?" Because we are bringing it out to the whole world and changing the game in how they consume their own data with these models.

>> Hillery, the other key piece of course, is developers. And so Red Hat obviously brings you a wealth of developer chops. So what we see is sometimes we call it super cloud, but it's a metaphor for a consistent experience irrespective of physical location. So you guys have this concept of hybrid by design. Some people talk about multi-cloud by design, they sort of go together and then that extends to hybrid AI. How do you think about that?

Varun Bijlani

>> Yeah, when we first started talking about hybrid cloud 2018 around the acquisition of Red Hat, a lot of it was about the developer. Giving the developer a consistent set of procedures and DevOps and tools and Kubernetes environment to operate in the same way across a bunch of different locations. And I think what's happening now is twofold. One, we have added on Instana, Turbonomic, Apptio, the ability to have observability, manage your workloads, optimize those workloads, optimize your costs across all environments in a consistent way, which gives you a platform that you're working with. And then in addition, AI is no longer something that's only in one place. The AI can be where your data is, where your workloads need to be, where your customers are. And so our platform approach to AI means then that both conversations, managing, viewing, optimizing a bunch of different environments and the developer experience can all be consistent. And then an AI platform that enables you to bring together the data needed for AI, those models, the governance of those models, monitoring of all of that, those can each be in any location across the hybrid cloud landscape, from on-premises to public cloud and out to the edge.

Varun Bijlani

>> I think I love what Hillery says here. And to just substantiate that further with the business priority lens, how does the CEO look at that same conversation? Because this conversation is no longer just a technology conversation.

>> Absolutely.

Varun Bijlani

>> And what we are also seeing is, earlier in that journey, there was such a rapid shift to let me get to the cloud and figure out how things would work, that we are seeing one in three programs, our CXOs are telling us that one in three of their programs are not on time or on budget. One in four are not really returning hard ROI. And when you start delving into why that's working, those who are successful, there are certain patterns emerging in that. Another call out five patterns that emerge. One, those that are successful are very product centric. And what I mean by that is business product centric. Like I have a loan origination product, I've got an insurance product, a baggage handling product. And they are looking at value in terms of those services that they provide their customers. That is determining the technical architectural choices that they're making. Earlier it was, "I'm going to go technology choices and I'm going to figure out how the business will consume it." Now we are seeing these are architectural choices depending on that business objective. And then to Hillery's point, this consistency of developer experience, consistency of operational experience is also becoming very, very important. That's why you're talking about this by design or this intentional thought process, which is now emerging.

>> And I look at those as somewhat data products. I mean, in many respects, the examples that you were giving. We see a couple of in terms of the whole hybrid cloud, hybrid AI. One is people went to the cloud and said, "Well, it's maybe a little too expensive. I can do it more cost-effectively on-prem." The other is very intentionally saying, "I'm going to run the experiment. And the POC and the MVP, I'm going to get it up and running in the cloud. And then specifically I have the intention of bringing it on-prem." Do you see those two vectors? And there's probably a bunch of stuff in between, but how do you see it?

Varun Bijlani

>> Absolutely, a hundred percent. A lot of people are looking to build and create to train their model or tune their model in a public cloud environment and then bring it on premises. We see clients even doing AI on the mainframe because they're wanting to do something at ultra-low latency. And so when latency comes into play, when cost, as Varun was alluding to, comes into play, you're going to choose to do the data preparation, the model creation, and then the model deployment in the locations that help you optimize all those factors.

>> I want to pick up on something Varun was talking about in terms of the ROI. You know, consumer AI, gen AI, the ROI of building million GPU clusters, I know we're not there yet, but it's pretty obvious. Better ads, you could get more revenue. I mean, okay, the ROI in the enterprise, a lot of singles being hit today and a lot of misses, swing and misses. To your point, a couple of data points, and I want you to see if it aligns with what you guys see. Our survey data with our partner, ETR, shows about 42% of the customers say they're funding gen AI from other budgets. And when you dig into those, it's non-IT budgets, it's business lines, it's productivity apps, some legacy RPA seems to be getting hit. And the ROI timelines are elongating because I think to your point, people are realizing, "Well, maybe it's not so easy." Because the apps today with GNI are very chatty. Okay. Are you seeing that, and how do you see that playing out over time?

Varun Bijlani

>> So I think spot-on in terms of it is singles being hit right now, and also this idea about tougher ROI. And I think one of the reasons of this tougher ROI was right now, it's a lot of experimentation. And they don't even know how to measure the ROI on the back of that. It's not all ROI in terms of revenue generation ROI, it's a lot of productivity use cases that are... And then how do you measure partial productivity? That becomes a whole new gamut of conversation. One of the examples, this client was looking at a large transformation program using cloud and AI. And when they first started looking at this, they said, "Our transformation is going to take us X number of years. ROI will start returning only in year four and five." And imagine taking that to the board. It's impossible. So then you start looking at saying, "Which business domain, what use case?" And you start driving ROI within the current year. And I think that's what we are seeing right now. And if you add to that, the conversation, if you've already heard about InstructLab.

>> Oh, amazing.

Varun Bijlani

>> Now that's magic in terms of ROI within the quarter, within the year. And when you start really extracting value of your own data, I think that's going to change the game significantly now.

>> What I see happening there, I love your thoughts, is you develop, you can train, that's great. Then you develop your rag, and then you can sort of automate the tuning, which is amazing. And you've got this open ecosystem. That will allow some ROI instantly that will throw off cash and then you can gain share. And then because people want some big, big hairy projects, but to your point, it's going to take four or five years to achieve those, and it's hard to get those funded. But it's going to take that long to solve some of the really challenging healthcare problems or logistical issues, etc. So what I would forecast is, you're going to see those singles, maybe turn it to doubles, throw off some cash. And then that will help fund or at least loosen up the purse strings. How do you think about that?

Varun Bijlani

>> Well, I see that the two conversations we started talking about hybrid and AI, sorry, cloud, let's say hybrid cloud and AI are together for a reason. We talked about why, but I think also from this, how do you fund AI perspective? The two are related. I mean, we see, we have one instance where in a cluster using Turbonomic, they've been able to reclaim 30% of the GPU utilization by doing appropriate workload balancing. By dropping down to that lower level of the IT stack and optimization, they just got back a of budget to drive their AI experimentation and things like that. Alternatively, we worked with the University of Queensland 74% faster in their medical research because of a storage upgrade to storage scale product. And so folks are looking at those lower levels to find optimization opportunities, find more throughput, and self-fund these AI projects by really eliminating technical debt, differences in operations across different environments. All those things we were talking about before can be a way that you can actually fund AI. It doesn't necessarily just have to be budget stealing from one category or the other.

Varun Bijlani

>> Oh, absolutely. Look at how core business processes are run, the payment process, the insurance handling process, the finance transaction process. What AI is now doing and the Gen AI is now doing in the conversation, it's not a bolt onto that particular process. Those processes are being completely rethought out. And when you start rethinking that entire process, you're getting 30 to 50% benefit because of the gen AI. You're getting up to 70% benefit because you've redefined that process. And I think that's what's going to start emerging as a big conversation. So going down to the lower levels to start exuding value and changing the way you execute on your processes.

>> And I guess I close saying, just making the observation that the cloud is no longer a remote set of services somewhere out there in the cloud. It's the public cloud, it's on-prem, it's out to the edge, and it's a substantially identical experience across with things like OpenShift. And that creates new opportunities. When you bring in AI. There may be some great tools in one cloud and you might have your own proprietary data that gives you advantage on-prem. You're going to use all those. But to the extent that you can create that seamless, sorry to use that word, but that abstraction that is going to create many, many new opportunities. I'll give you the last word. I know you got a hard stop, so please, Hillery, give us the close.

>> Well, I'll just say you want to get a grip on your data, right? That's why the AI platform conversation for us, starts with data and data platforms. So you get a grip on your data across that landscape. That's the first step. Then you have your AI capabilities. And then at the back end, so to say, you have the governance in the monitoring of those AI capabilities, no matter where they are. And if you think of this whole environment holistically and look at the underlying infrastructure and the opportunities to create consistency across environments, you can get AI at scale safely and in almost a self-funding kind of a way. And I think that's the really exciting opportunity of how these two conversations are coming together.

>> Love it. Final thoughts, Varun.

Varun Bijlani

>> I'll just add on to that, actually. And in order to help our clients execute on this, how do you go about doing what Hillery just said? And our hybrid by design framework takes you on that journey. How do you start looking at all of these different aspects, right? From which business products to what kind of teams, what kind of architecture, the gen AI platforms, the security requirements. The framework allows you to demystify that journey and actually start releasing value. So, hybrid by design comes from there.

>> Guys, I posted on my LinkedIn. I haven't been this excited about IBM in 10 years. I meant it and I can tell that you guys are excited as well. Really appreciate your time.

>> Thanks so much.

Varun Bijlani

>> Thank you so much.

>> Okay. Keep it right there. We're not live, but we will be shortly on demand at theCUBE.net. This is Dave Vellante for theCUBE. We'll be right back.

Varun Bijlani & Hillery Hunter | IBM

search

>> Hi, everybody. Welcome to the special CUBE presentation. We're here at the BECC in Boston where theCUBE started in 2010, the first CUBE ever at EMC World in 2010. Really excited to have this special presentation here. Inside the IBM studio, it's fantastic. Varun Bijlani is here. He's the Global Managing Partner for Hybrid Cloud Services at IBM. And Hilary Hunter, who's the CTO and GM Innovation, love that title, for IBM Infrastructure. Folks, welcome back to theCUBE-ish.

Varun Bijlani

>> Thanks so much.

Varun Bijlani

>> Thank you so much.

>> Thank you.

>> Great to see you guys. So last week we put out a post on our Breaking Analysis, talking about the sort of intersection between hybrid cloud and AI, what we called hybrid AI. We clearly see these two things coming together in a different pace. And so, I wonder if you could share your thoughts on what you're seeing from customers going around this show at IBM think. What are you hearing? What are you seeing? What do you expect for the intersection of those two concepts?

>> Yeah, I definitely want to hear a little bit more about your data. But I think just to share some of ours, if you look back like six or seven years ago, hybrid cloud wasn't so much a thing. Cloud was sort of a place where you went to do your computing, where you went to use your AI. And I think the reality from our polling is, today, if you ask CIOs how much their cloud estate represents hybrid cloud, about 70% of them will say that they actually have a hybrid cloud infrastructure. So hybrid cloud is a thing, it's a thing everyone is doing. But it often when you dig into the details, is actually a bunch of disparate and heterogeneous cloud environments, and not a unified place that you really have a platform to create new things.

>> So, thank you for that. Varun, I think in some ways, IBM maybe has an unfair bias, and here's what I mean by that. What our data shows is that, and I think this has become sort of a bromide, bringing AI to the data and a lot of data on-prem. I think a lot tons of enterprise data, if you take out video for consumers is on-prem. What I mean by the unfair bias is in order for hybrid AI to happen, you've got to have the tooling, you've got to have the LLM optionality. You've got to have the ecosystem. Actually, IBM has those things. That's why I say it's sort of an unfair bias. A lot of the sort of traditional on-prem vendors maybe don't have that. I wonder if you could comment on how you're bringing that all together.

Varun Bijlani

>> I think the first fundamental thought out here is that there is a very significant symbiotic relationship between hybrid cloud and AI. And today, AI is supremely accelerating the execution of hybrid cloud. And at the same time, you can't really go from pilot to production to scale without a robust architecture. So, I think that's the first reason why this conversation right now is so relevant. And then in terms of our own independence, our thought process around this is about making sure that we are allowing you to accelerate the value of your own data. Up until now, all the conversations have been about, I've got models with large amount of public data, but the amount of enterprise data in there is next to negligible. Your differentiation is going to come when you can take your data and use that with those right models. And I think that's where we've decided to make our conversation much more open. And you heard today, Granite is now open-sourced. And then you start saying, "Where do we have the unfair advantage?" Because we are bringing it out to the whole world and changing the game in how they consume their own data with these models.

>> Hillery, the other key piece of course, is developers. And so Red Hat obviously brings you a wealth of developer chops. So what we see is sometimes we call it super cloud, but it's a metaphor for a consistent experience irrespective of physical location. So you guys have this concept of hybrid by design. Some people talk about multi-cloud by design, they sort of go together and then that extends to hybrid AI. How do you think about that?

Varun Bijlani

>> Yeah, when we first started talking about hybrid cloud 2018 around the acquisition of Red Hat, a lot of it was about the developer. Giving the developer a consistent set of procedures and DevOps and tools and Kubernetes environment to operate in the same way across a bunch of different locations. And I think what's happening now is twofold. One, we have added on Instana, Turbonomic, Apptio, the ability to have observability, manage your workloads, optimize those workloads, optimize your costs across all environments in a consistent way, which gives you a platform that you're working with. And then in addition, AI is no longer something that's only in one place. The AI can be where your data is, where your workloads need to be, where your customers are. And so our platform approach to AI means then that both conversations, managing, viewing, optimizing a bunch of different environments and the developer experience can all be consistent. And then an AI platform that enables you to bring together the data needed for AI, those models, the governance of those models, monitoring of all of that, those can each be in any location across the hybrid cloud landscape, from on-premises to public cloud and out to the edge.

Varun Bijlani

>> I think I love what Hillery says here. And to just substantiate that further with the business priority lens, how does the CEO look at that same conversation? Because this conversation is no longer just a technology conversation.

>> Absolutely.

Varun Bijlani

>> And what we are also seeing is, earlier in that journey, there was such a rapid shift to let me get to the cloud and figure out how things would work, that we are seeing one in three programs, our CXOs are telling us that one in three of their programs are not on time or on budget. One in four are not really returning hard ROI. And when you start delving into why that's working, those who are successful, there are certain patterns emerging in that. Another call out five patterns that emerge. One, those that are successful are very product centric. And what I mean by that is business product centric. Like I have a loan origination product, I've got an insurance product, a baggage handling product. And they are looking at value in terms of those services that they provide their customers. That is determining the technical architectural choices that they're making. Earlier it was, "I'm going to go technology choices and I'm going to figure out how the business will consume it." Now we are seeing these are architectural choices depending on that business objective. And then to Hillery's point, this consistency of developer experience, consistency of operational experience is also becoming very, very important. That's why you're talking about this by design or this intentional thought process, which is now emerging.

>> And I look at those as somewhat data products. I mean, in many respects, the examples that you were giving. We see a couple of in terms of the whole hybrid cloud, hybrid AI. One is people went to the cloud and said, "Well, it's maybe a little too expensive. I can do it more cost-effectively on-prem." The other is very intentionally saying, "I'm going to run the experiment. And the POC and the MVP, I'm going to get it up and running in the cloud. And then specifically I have the intention of bringing it on-prem." Do you see those two vectors? And there's probably a bunch of stuff in between, but how do you see it?

Varun Bijlani

>> Absolutely, a hundred percent. A lot of people are looking to build and create to train their model or tune their model in a public cloud environment and then bring it on premises. We see clients even doing AI on the mainframe because they're wanting to do something at ultra-low latency. And so when latency comes into play, when cost, as Varun was alluding to, comes into play, you're going to choose to do the data preparation, the model creation, and then the model deployment in the locations that help you optimize all those factors.

>> I want to pick up on something Varun was talking about in terms of the ROI. You know, consumer AI, gen AI, the ROI of building million GPU clusters, I know we're not there yet, but it's pretty obvious. Better ads, you could get more revenue. I mean, okay, the ROI in the enterprise, a lot of singles being hit today and a lot of misses, swing and misses. To your point, a couple of data points, and I want you to see if it aligns with what you guys see. Our survey data with our partner, ETR, shows about 42% of the customers say they're funding gen AI from other budgets. And when you dig into those, it's non-IT budgets, it's business lines, it's productivity apps, some legacy RPA seems to be getting hit. And the ROI timelines are elongating because I think to your point, people are realizing, "Well, maybe it's not so easy." Because the apps today with GNI are very chatty. Okay. Are you seeing that, and how do you see that playing out over time?

Varun Bijlani

>> So I think spot-on in terms of it is singles being hit right now, and also this idea about tougher ROI. And I think one of the reasons of this tougher ROI was right now, it's a lot of experimentation. And they don't even know how to measure the ROI on the back of that. It's not all ROI in terms of revenue generation ROI, it's a lot of productivity use cases that are... And then how do you measure partial productivity? That becomes a whole new gamut of conversation. One of the examples, this client was looking at a large transformation program using cloud and AI. And when they first started looking at this, they said, "Our transformation is going to take us X number of years. ROI will start returning only in year four and five." And imagine taking that to the board. It's impossible. So then you start looking at saying, "Which business domain, what use case?" And you start driving ROI within the current year. And I think that's what we are seeing right now. And if you add to that, the conversation, if you've already heard about InstructLab.

>> Oh, amazing.

Varun Bijlani

>> Now that's magic in terms of ROI within the quarter, within the year. And when you start really extracting value of your own data, I think that's going to change the game significantly now.

>> What I see happening there, I love your thoughts, is you develop, you can train, that's great. Then you develop your rag, and then you can sort of automate the tuning, which is amazing. And you've got this open ecosystem. That will allow some ROI instantly that will throw off cash and then you can gain share. And then because people want some big, big hairy projects, but to your point, it's going to take four or five years to achieve those, and it's hard to get those funded. But it's going to take that long to solve some of the really challenging healthcare problems or logistical issues, etc. So what I would forecast is, you're going to see those singles, maybe turn it to doubles, throw off some cash. And then that will help fund or at least loosen up the purse strings. How do you think about that?

Varun Bijlani

>> Well, I see that the two conversations we started talking about hybrid and AI, sorry, cloud, let's say hybrid cloud and AI are together for a reason. We talked about why, but I think also from this, how do you fund AI perspective? The two are related. I mean, we see, we have one instance where in a cluster using Turbonomic, they've been able to reclaim 30% of the GPU utilization by doing appropriate workload balancing. By dropping down to that lower level of the IT stack and optimization, they just got back a of budget to drive their AI experimentation and things like that. Alternatively, we worked with the University of Queensland 74% faster in their medical research because of a storage upgrade to storage scale product. And so folks are looking at those lower levels to find optimization opportunities, find more throughput, and self-fund these AI projects by really eliminating technical debt, differences in operations across different environments. All those things we were talking about before can be a way that you can actually fund AI. It doesn't necessarily just have to be budget stealing from one category or the other.

Varun Bijlani

>> Oh, absolutely. Look at how core business processes are run, the payment process, the insurance handling process, the finance transaction process. What AI is now doing and the Gen AI is now doing in the conversation, it's not a bolt onto that particular process. Those processes are being completely rethought out. And when you start rethinking that entire process, you're getting 30 to 50% benefit because of the gen AI. You're getting up to 70% benefit because you've redefined that process. And I think that's what's going to start emerging as a big conversation. So going down to the lower levels to start exuding value and changing the way you execute on your processes.

>> And I guess I close saying, just making the observation that the cloud is no longer a remote set of services somewhere out there in the cloud. It's the public cloud, it's on-prem, it's out to the edge, and it's a substantially identical experience across with things like OpenShift. And that creates new opportunities. When you bring in AI. There may be some great tools in one cloud and you might have your own proprietary data that gives you advantage on-prem. You're going to use all those. But to the extent that you can create that seamless, sorry to use that word, but that abstraction that is going to create many, many new opportunities. I'll give you the last word. I know you got a hard stop, so please, Hillery, give us the close.

>> Well, I'll just say you want to get a grip on your data, right? That's why the AI platform conversation for us, starts with data and data platforms. So you get a grip on your data across that landscape. That's the first step. Then you have your AI capabilities. And then at the back end, so to say, you have the governance in the monitoring of those AI capabilities, no matter where they are. And if you think of this whole environment holistically and look at the underlying infrastructure and the opportunities to create consistency across environments, you can get AI at scale safely and in almost a self-funding kind of a way. And I think that's the really exciting opportunity of how these two conversations are coming together.

>> Love it. Final thoughts, Varun.

Varun Bijlani

>> I'll just add on to that, actually. And in order to help our clients execute on this, how do you go about doing what Hillery just said? And our hybrid by design framework takes you on that journey. How do you start looking at all of these different aspects, right? From which business products to what kind of teams, what kind of architecture, the gen AI platforms, the security requirements. The framework allows you to demystify that journey and actually start releasing value. So, hybrid by design comes from there.

>> Guys, I posted on my LinkedIn. I haven't been this excited about IBM in 10 years. I meant it and I can tell that you guys are excited as well. Really appreciate your time.

>> Thanks so much.

Varun Bijlani

>> Thank you so much.

>> Okay. Keep it right there. We're not live, but we will be shortly on demand at theCUBE.net. This is Dave Vellante for theCUBE. We'll be right back.