(upbeat music) >> Hello everyone and welcome to theCUBE. In this special power panel we're going to dig into and take a peek at the future of cloud. You know a lot has transpired in the last decade. The cloud itself, we've seen a data explosion. The AI winter turned into machine intelligence going mainstream. We've seen the emergence of As-a-Service models. And as we look forward to the next 10 years we see the whole idea of cloud expanding, new definitions occurring. Yes, the world is hybrid but the situation is more nuanced than that. You've got remote locations, smaller data centers, clandestine facilities, oil rigs, autonomous vehicles, windmills, you name it. Technology is connecting our world, data is flowing through the pipes like water, and AI is helping us make sense of the noise. All of this, and more is driving a new digital economy. And with me to talk about these topics are three great guests from IBM. Maia Sisk is the Director of SaaS Offering Management, at IBM Data and AI. And she's within the IBM Cloud and Cognitive Software Group. Stephanie Walter is the Program Director for data and AI Offering Management, same group IBM Cloud and Cognitive Software. And Daniel Berg is a Distinguished Engineer. He's focused on IBM Cloud Kubernetes Service. He's in the Cloud Organization. And he's going to talk today a lot about IBM's cloud Satellite and of course Containers. Wow, two girls, two boys on a panel, we did it. Folks welcome to theCUBE. (chuckles) >> Thank you. >> Thank you. >> Glad to be here. >> So Maia, I want to start with you and have some other folks chime in here. And really want to dig into the problem statement and what you're seeing with customers and you know, what are some of the challenges that you're hearing from customers? >> Yeah, I think a big challenge that we face is, (indistinct) talked about it earlier just data is everywhere. And when we look at opportunities to apply the cloud and apply an As-a-Service model, one of the challenges that we typically face is that the data isn't all nice cleanly package where you can bring it all together, and you know, one AI models on it, run analytics on it, get it in an easy and clean way. It's messy. And what we're finding is that customers are challenged with the problem of having to bring all of the data together on a single cloud in order to leverage it. So we're now looking at IBM and how we flip that paradigm around. And instead of bringing the data to the cloud bring the cloud to the data , in order to help clients manage that challenge and really harness the value of the data, regardless of where you live. >> I love that because data is distributed by its very nature it's silo, Daniel, anything you'd add? >> Yeah, I mean, I would definitely echo that, what Maia was saying, because we're seeing this with a number of customers that they have certain amount of data that while they're strategically looking that moving to the cloud, there's data that for various reasons they can not move itself into the cloud. And in order to reduce latency and get the fastest amount of processing time, they going to move the processing closer to that data. And that's something that we're looking at providing for our customers as well. The other services within IBM Cloud, through our notion of IBM Cloud Satellite. How to help teams and organizations get processing power manage them to service, but closer to where their data may reside. >> And just to play off of that with one other comment. Then the other thing I think we see a lot today is heightened concerned about risks, about data security, about data privacy. And you're trying to figure out how to manage that challenge of especially when you start sending data over the wire, wanting to make sure that it is still safe, it is still secure and it is still resident in the appropriate places. And that kind of need to manage the governance of the data kind of adds an additional layer of complexity. >> Right, if it's not secure, it's a, non-starter, Stephanie let's bring you into the conversation and talk about, you know, some of the waves that you're seeing. Maybe some of the trends, we've certainly seen digital accelerate as a result of the pandemic. It's no longer I'll get to that someday. It's really, it become a mandate you're out of business, if you don't have a digital business. What are some of the markets shifts that you're seeing? >> Well, I mean, really at the end of the day our clients want to infuse AI into their organizations. And so, you know, really the goal is to achieve ambient AI, AI that's just running in the background unchoosibly helping our clients make these really important business decisions. They're also really focused on trust. That's a big issue here. They're really focused on, you know, being able to explain how their AI is making these decisions and also being able to feel confident that they're not introducing harmful biases into their decision-making. So I say that because when you think about, you know digital organization going digital, that's what our customers want to focus on. They don't want to focus on managing IT. They don't want to focus on managing software. They don't want to to have to focus on, you know, patching and upgrading. And so we're seeing more of a move to manage services As-a-Service technologies, where the clients can really focus on their business problems and using The technologies like AI, to help improve their businesses. And not have to worry so much about building them from the ground up. >> So let's stay on that for a minute. And maybe Maia, Daniel, you can comment. So you, Stephanie, you said that customers want to infuse AI and kind of gave some reasons why, but I want to stay on that for a minute. That, what is that really that main outcome that they're looking for? Maybe there are several, they're trying to get to insight. You mentioned that trynna be more efficient it sounds like they're trynna automate governance and compliance, Maia, Daniel can you sort of add anything to this conversation? >> Yeah, well, I would, I would definitely say that, you know at the end of the day, customers are looking to use the data that they have to make smarter decisions. And in order to make smarter decisions it's not enough to just have the insight. The insight has to, you know, meet the business person that needs it, you know in the context, you know, in the application, in the customer interaction. So I think that that's really important. And then everything else becomes like the the superstructure that helps power, that decision and the decision being embedded in the business process. So we at IBM talk a lot about a concept we call the Ladder to AI. And the the short tagline is there is no AI without IA. You know, there is no Artificial Intelligence without Information Architecture. It is so critical, you know, Maia's version this is the garbage in garbage out. You have to have high quality data. You have to have that data be well-organized and well-managed so that you're using it appropriately. And all of that is just, you know then becomes the fuel that powers your AI. But if you have the AI without having that super structure, you know, you're going to end up making, get bad decisions. And ultimately, you know our customers making their customers experience less than it could and should be. And in a digital world, that's, you know, at the end of the day, it's all about digitizing that interaction with whoever the end customer whoever the end consumer is and making that experience the best it can be, because that's what fuels innovation and growth. >> Okay. So we've heard Arvind Krishna talk about, he actually made this statement IBM has to win the architectural battle for cloud. And I'm wondering maybe Daniel you can comment, on what that architectural framework looks like. I mean Maia just talked about the Information Architecture. You can't have AI without that foundation but we know what does Arvind mean by that? How is IBM thinking about that? >> Yeah, I mean, this is where we're really striving to allow our customers really focusing on their business and focusing on the goals that they're trying to achieve without forcing them to worry as much about the IT and the infrastructure and the platform for which they're going to run. Typically, if you're anchored by your data and the data is not able to move into the cloud, generally we would say that you don't have access to cloud services. You must go and install and run and operate your own software to perform the duties or the processing that you require. And that's a huge burden to push onto a customer because they couldn't move their data to your cloud. Now you're pushing a lot of responsibilities back onto them. So what we're really striving for here is, how can we give them that cloud experience where they can process their data? They can run their run book. They can have all of that managed As-a-Service so that they could focus on their business but get that closer to where the data actually resides. And that's what we're really striving for as far as the architecture is concerned. So with IBM Cloud Satellite, we're pushing the core platform and the platform services that we support in IBM Cloud outside of our data centers and into locations where it's closer to your data. And all of that is underpinned by Containerizations, Containers, Kubernetes and OpenShift. Is fundamentally the platform for which we're building upon. >> Okay. So that, so really it's still it's always a data problem, right? Data is you don't want to move it if you don't have to. Right. So it's, so Stephanie, should we think about this as a new emergent data architecture I guess that's what IA is all about. How do you see that evolving? >> Well I mean, I see it evolving as, I mean, first of all our clients, you know, we know that data is the lifeblood of AI. We know the vast majority of our clients are using more than one cloud. And we know that the client's data may be located in different clouds, and that could be due to costs, that could be due to location. So we have to ask the question, how are our clients supposed to deal with this? This is incredibly complex environments they're are incredibly complex reasons sometimes for the data to be where it is. It can include anything from costs to laws, that our clients have to abide by. So what we need to do, is we need to adapt to these different environments and provide clients with the consistent experience and lower complexity to be able to handle data and be able to use AI in these complex environments. And so, you know, we know data, we also know data science talent is scarce. And if each one of these environments have their own tools that need to be used, depending on where the data is located, that's a huge time sink, for these data scientist and our clients don't want to waste their talents time on problems like this. So what we're seeing is, we're seeing more of a acceptance and realization that this is what our clients are dealing with. We have to make it easier. We have to do Innovative things like figure out how to bring the AI to the data, how to bring the AI to where the clients need it and make it much easier and accessible for them to take advantage of. >> And I think there's an additional point to make on this one, which is it's not just easy and accessible but it's also unified. I mean, one of the challenges that customers face in this multicloud environment and many customers are multicloud, you know, not necessarily by intent but just because of how, you know, businesses have adopted as a service. But to then have all of that experience be fragmented and have different tools not just of data, but different pools of, again catalog, different pools of data science it's extremely complex to manage. So I think one of the powerful things that we're doing here, is we're kind of bringing those multiple clouds together, into more of an integrated or a unified, you know window into the client's data in AI state. So not only does the end-user not have to worry about you know, the technologies of dealing with multiple individual clouds, but also, you know it all comes together in one place. So it can be give managed in a more unified way so that assets can be shared across, and it becomes more of a unified approach. The way I like to think of it is, you know, it's true hybrid multicloud, in that it is all connected as opposed to multi-cloud, but it's pools of multiple clouds, one cloud at a time. >> So it can we stay on that for a second because it's, you're saying it's unified but the data stays where it is. The data is distributed by nature. So it's unified logically, but it's decentralized. Is that, am I getting that right? Correct. Okay. Correct. All right. I'm really interested in how you do this. And maybe we can talk about maybe the approach that you take for some of your offerings and maybe get specific on that. So maybe Stephanie, why don't you start, you know, Yes so, what do you have in your basket? Like Cloud Pak So what we have in our basket I mean lets talk about that. >> We have, so Cloud Pak for Data as a Service. This is our premier data and AI platform. It's offered as a service, its fully managed, and there's roughly, there's 30 services integrated services in our services catalog and growing. So we have services to help you through the entire AI life cycle from preparing your data, which is Maia was saying it's very, very, very important. It's critical to any successful AI project. From building your models, from running the models and then monitoring them to make sure that as I was saying before, you can trust them. You don't have to make sure that, you need to make sure that there's not biased. You need to be able to manage these models and then the life cycle them retrain them if needed. So our platform handles all of that. It's hosted on IBM Cloud. And what we're doing now, which is really exciting, is we're going to use, and you mentioned before IBM Cloud Satellite, as a way for us to send our AI to data that perhaps is located on another cloud or another environment. So how this would work is that the services that are integrated with Cloud Pak for Data as a Service they'll be able to use satellite locations to send their AI workloads, to run next to the data. And this means that the data doesn't need to be moved. You don't have to worry about high egress charges. You can see, you can reduce latency and see much stronger performance by running these AI workloads where it counts. We're really excited to to add this capability to our platform. Because, you know, we spent a lot of time talking about earlier all of these challenges that our clients have and this is going to make a big difference in helping them overcome them. Okay. So Daniel, how to Containers fit in? I mean, obviously the Red Hat acquisition was so strategic. We're seeing the real, the ascendancy of OpenShift in particular. Talk about Containers and where it fits into the IBM Cloud Satellite strategy. >> Yeah. So a lot of this builds on top of how we run our cloud business today. Today the vast majority of the services that are available in IBM cloud catalog, actually runs as Containers, runs in a Kubernetes based environment and runs on top of the services that we provide to our customers. So the Container Platform that we provide to our customers is the same one that we're using to run our own cloud services. And those are underpinned with Containers, Kubernetes, and OpenShift. And IBM cloud satellite, based on the way that the designed our Container Platform using Kubernetes and Containers and OpenShift, allows us to take that same design and the same principles and extended outside of our data centers with user provided infrastructure. And this, this goes back to what Stephanie was saying is a satellite location. So using that technology, that same technology and the fact that we've already containerized many of our services and run them on our own platform, we are now distributing our platform outside of IBM Cloud Data Centers using satellite locations and making those available for our cloud service teams, to make their services available in those locations. >> I see and Maia, this, it is as a service. It's a OPEX. Is that right? Absolutely Okay. Absolutely >> Yeah, it's with the two different options on how we can run. One is we can leverage IBM Cloud Satellite and reach into a customer's operating environment. They provide the infrastructure, but we've provide the As-a-Service experience for the Container on up. The other option that we have is for some of our capabilities like our data science capability, where, you know customer might need something a little bit more turnkey because it's, you know, more of a business person or somebody in the CTO's office consuming the As-a-Service. We'll also offer select workloads in an IBM own satellite and environment. I, you know, so that it kind of soup to nuts managed by us. But that is the key is that other than, you know providing the operating environment and then connecting what we do to, you know, their data sources, really the rest is up to us. We're responsible for, you know everything that you would expect in an As-a-Service environment. That things are running, that they're updated, that they're secure, that they're compliant, that's all part of our responsibility. >> Yeah. So a lot of options for customers and it's kind of the way they want to consume. Let's talk about the business impact. You know, you guys, IBM, very consultative selling, you know, tight relationships with customers. What's the business case look like when you go into a client? What's the conversation like? What's possible? What can you share? Stephanie, can you maybe start things off there? Any examples, use-cases, business case, help us understand the metrics. >> Yeah. I mean, so let's talk about a couple of use cases here. So let's say I'm an investment firm, and I'm using data points from all kinds of data sources right? To use AI, to create models to inform my investment decisions. So I'm going to be using, I may be using data sources you know, like regulatory filings, newspaper articles that are pretty standard. I may also be using things like satellite data that monitors parking lots or maybe even weather data, weather forecast data. And all of this data is coming together and being, it needs to be used for models to predict, you know when to buy, sell, trade, however, due to costs, due to just availability of the data they may be located on completely different clouds. You know, and we know that especially capital markets things are fast, fast, fast. So I need to bring my AI to my data, and need to do it quickly so that I can build these models where the data resides, and then be able to make my investment decisions, very fast. And these models get updated often because conditions change, markets change. And this is one way to provide a unified set of AI tools that my data scientists can use. We don't have to be trained on I'm told depending on what cloud the data is stored on. And they can actually build these models much faster and even cheaper. If you would take into egress charges into consideration, you know, moving all the all this data around. Another use case that we're seeing is you know, something like let's say, a multinational telecommunications company that has locations in multiple countries and maybe they want to reduce their customer churn. So they have say customer data that it's stored in different countries and different countries may have different regulations, or the company may have policies that, that data can't be moved out to those country. So what can we do? Again, what we can do is we can send our AI to this data. We can make a customer churn prediction model, that when my customer service representative is on the phone with a customer, and put their information, and see how likely they are to stop using my service and tailor my phone interaction and the offers that I would offer them as this customer service representative to them. If there's a high likelihood that they're going to churn I will probably sweeten the deal. And I can do all that while I'm being fast, right. Because we know that these interactions need to happen quickly. But also while complying with whatever policies or even regulations that are in place for my multinational company. So you know, if you think back to the use cases that I was just talking about you know, latency, performance, reducing costs and also being able to comply with any policy or regulations that our customers might have are really, are really the key pieces of the use cases that we've been seeing. >> Yeah. So Maia there's a theme here. I bring five megabytes of code to a petabyte of data kind of thing. And so Stephanie was talking about speed. There's a an inherent compliance and governance piece. It's it sounds like it's not a bolt on, it's not an afterthought, it's fundamental. So maybe you could add to the conversation, just specifically interested in, you know, what should a client expect? I mean, you're putting data in the hands of you know domain experts in the line of business. There's a self-serve component here, presumably. So there's cross selling is what I heard in some of what Stephanie was just talking about. So it was revenue, there's cost cutting, there's risk reduction, that I'm seeing the business case form. What can you add? >> Yeah, absolutely. I think that the only other thing I would add, is going back to the conversation that we had about, Oh you know, a lot of this is being driven by, you know the digitization of business and you know even moreso this year. You know, at the end of the day there's a lot of costs benefits to leveraging and As-a-Service model, you know, to leveraging that experience in economies of scale from a service provider, as well as, you know leveraging satellite kind of takes that to the next level of, you know, reducing some other costs. But I always go back to, you know at the end of the day, this is about customer experience. It's about revenue creation, and it's about, you know, creating, you know enhanced customer satisfaction and loyalty. So there's a top-line benefits here, you know, of having the best possible AI, you know plugging that into the customer experience, the application where that application resides. So it's not just about where the data resides. You can also put it on the other side and say, you know, we're bringing the AI, we're bringing the machine learning model to the application so that the experiences at excellent the application is responsive there's less latency and that can help clients then leverage AI to create those revenue benefits, you know, of having the the satisfied customer and of having the, you know the right decision at the right time in order to, you know propel them to, to spend and spend more. >> So Daniel bring us home. I mean, there's a lot of engineering going on here. There's the technology, the people in the process if I'm a client, I'm going to say, okay, I'm going to rely on IBM R&D to cut my labor costs, to drive automation, to help me, you know, automate governance and reduce my risks, you know, take care of the technology. You know, I'll focus my efforts on my process, my people but it's a journey. So how do you see that shaping out in the next, you know several years or, or the coming decade, bring us home. >> Yeah. I mean what we're seeing here is that there's a realization that customers have highly skilled individuals. And we're not saying that these highly skilled individuals couldn't run and operate these platforms and the software themselves, they absolutely could. In some cases, maybe they can't but in many cases they could. But we're also talking about these are they're highly skilled individuals that are focusing on platform and platform services and not their business. And the realization here is that companies want their best and brightest focused on their business, not the platform. If they can get that platform from another vendor that they rely on and can provide the necessary compute services, in a timely and available fashion. The other aspect of this is, people have grown to appreciate those cloud services. They like that on demand experience. And they want that in almost every aspect of what they're working on. And the problem is, sometimes you have to have that experience in localities that are remote. They're very difficult. There's no cloud in some of these remote parts of the world. You might think that clouds everywhere, but it's not. It's actually in very specific locations across the world, but there are many remote locations that they want and need these services from the cloud that they can get. Something like IBM Cloud Satellite. That is what we're pursuing here, is being able to bring that cloud experience into these remote locations where you can't get it today. And that's where you can run your AI workloads. You don't have to run it yourself, we will run it and you can put it in those remote locations. And remote locations don't actually have to be like in the middle of a jungle, they could be in your, on your plant floor or within a port that you have across the world, right? It could be in a warehouse. I mean, there's lots of areas where there's data that needs to be processed quickly, and you want to have that cloud experience, that usage pay model for that processing. And that's exactly what we're trying to achieve with IBM Cloud Satellite and what we're trying to achieve with the IBM Cloud Pak for Data as a Service as well. Running on satellite is to give you those cloud experiences. Those services managed as a service in those remote locations that you absolutely need them and want them. >> Well, you guys are making a lot of progress in the next decade is not going to look like the last decade. I can pretty confident in that prediction. Guys thanks so much for coming on the cube and sharing your insights, really great conversation. >> Absolutely. Thank you, Dave. >> Thank you. >> You're welcome, and thank you for watching everybody. This is Dave Vellante from the cube. We'll see you next time. (upbeat music)