Defining Your Cloud Core Migration Strategy

Hello, silver liners, and welcome back to the telco core strategy summit. I'm Liz Coin managing editor here at silver linings. And I'm thrilled that you joined us again. Before we start, our second session, I wanna thank our sponsors for today's event, Kentik and Rakuten Symphony. Both of them have loaded great content in our content hub on the top left hand side of your screen, so be sure to check that out. Before we get started on the second session, a few housekeeping reminders, Q and A tab is located on the top left hand side of your screen. If you're having IT issues, you can simply refresh your browser, and all content will be available on demand. We've also got some social media channels for you You can join us on silver linings on LinkedIn and go silver linings on Twitter or x and the hashtag is telco core summit. Let's talk about session two. The title of session is defining your cloud core migration strategy fail to prepare and prepare to fail. We're kicking it off with a great keynote from Colin Bannon of BT business. Then we're moving into a panel with orange business US cellular Rakuten Symphony, Kentik, and it's moderated by Monica Powell from Sanzaphili. It's gonna be a great panel. And then we've got a closing keynote from our good friend Eduardo Ponce at Telecom Argentina. So with that, I'm gonna hand it over to Colin for his keynote. Hello, everybody. Lovely to see you. I'm gonna spend about fifteen minutes right now to go through, you know, what I would call planning to win some considerations that we've learned on our transformation of our core to cloud. So I wanna just maybe start with a bit of a, a history lesson and tell you a story, you know, if you think about, For many of you, you're familiar with this, but there's a few pitfalls and learnings that we've had along the way. But to start that story, it would go back a few years. Maybe five to ten years. When we looked at, you know, this first trend, the journey to cloud where, We found that many, of ourselves, you, yourselves, and your leadership team started to develop that modern cloud based microservices front end that was for your b to c or your b to b, and that went down with often really great success. I mean, our leadership teams, you know, on the back of that said, well, hey, let's that sounds like a really great idea. Let's Let's let's, you know, let's move everything to cloud. And often that was done without the business cases behind it or understanding what the challenges were, but you would have heard the term a few years back cloud first. Now, obviously, you you went through that And and I would call that, what I would call that is the, the great re platforming. And to some extent, we're still going through that. But what we found in the in that journey was a set of roadblocks. There was clearly a lack of development skills. Significant, issues around as we started on this application performance and resilience. You know, if you're think about the the the best practices of what you had in the monolithic stack in the data center, moving that to cloud, the first, ventures into that. There was real challenges on how you would have the same discipline and the same diversity and and resilience when you start to network out into the cloud. Obviously, as different teams started to move their workloads to cloud and move their core aspects as well, you found that there was different models and different cloud providers as well. You know, multi cloud, you get fragmented operational security. And then we would start to find the things that are getting more and more significant regulations, data sovereignty requirements around how your, your data is handled and, what CSPs are managing that data outside of the and the regulations that impact. Today, mostly around the world, it's about regulations, data sovereignty, is regulations around data at rest, I e, where does that hard drive sit? Whereas the reality is we're starting to see also regulations and the regulators starting to move on to, talking about data in motion as well. So that's the the application flows between the the, the hard drive and the users and the applications. And then the last one, probably was just the sheer inertia and volume and size and the weight of the that ice varague of enterprise apps that were on the legacy platforms. And what we found very quickly on the back of that, it was we went sort of through this sort of hype cycle of cloud first into essentially where most of us are today, which is cloud complexity. And we needed a way to move on from that to really start unlocking the cloud. So, our point of view, BT's point of view is the reality is the there is a lot of hype and a lot of friction and, stories around, you know, private versus public. And we really, wanna take a step back from that and and and really take the point of view is to actually, we take a point of view where is let the app decide. And there's a set of criteria around apps, not just latency, and not just resilience in use case, but actually different locations, attract different economic models as well. So there is a commercial aspect that people are starting to light up. You often hear terms around bill shock, egress charges and others. Those are the sort of things that people are lighting up to say, well, some apps have a Pensity to be in public cloud, accessed via Internet. Other apps have a propensity for private. Some apps, of course. And if you look at the the, the slide that I have up here, if you think about it as the solar system, You know, the closer you get to the customer edge, probably the more expensive compute you'll get, the best economies of scales are further out, from that. This is not strictly speaking a latency, though there is some rough analogy around latency. So what we're starting to see is clusters of use cases. And and also this drives other implications of how you need to abstract and actually have a cloud operating system. And to be able to, to support distributed workloads and your workloads and the applications that you develop may need to be in multiple locations, depending on your product, your your business requirements, and your customer's requirements. The customer's edge, as I said, Often, it is the most expensive type of compute. But if you have the right business cases, you know, there are some as, you know, many of the experts out there talk about where you need to have that on prem analytics, that on prem processing, aspect. We're starting to see other sort of rings, and you can slice these next sort of rings up into a couple of different areas. IoT to the rugged edge, the mobile edge, Then the CNF Edge, which others call it the, how would you say the, the carrier hotels, those hyper interconnected locations where five g mobile providers, the ISPs appearing, the public cloud and the SaaS providers are exposing their service provider edge. Now that's also very, very expensive sort of rent for your workloads. However, it's a great place to insert security insertion layers, service insertion layers for network virtual functions as well there because That's a great choke point for where all the the this thing called cloud or the the this thing called networks and the internet They all interconverge there. It's a great place for network functions. Maybe not the greatest place for apps. And then the ones that we all know and love, data centers, private cloud edge, and then public cloud, and you got multi cloud. And then, of course, with other services, you're starting to see compute even further out in the far future up in satellite when you start to do, QKD and some of the other services, and you look at some of the Apple releases recently. There's also distributed compute, starting to happen up there. So, really, the future is distributed in our view. And, even today, you see, work cases where it's also hybrid. So you've got maybe still a mainframe registering the journal, on the backend with some middleware, then you've got, you know, third party services on private cloud edge, and then you've got your modern microservices front ends on the public cloud edge or some of your service functions there. The reality is what it means is that the wide area network is really performing the function of what used to be the data center LAN. And creating a fabric out of that area is one of our beliefs, and one of our learnings about moving, moving your core to the cloud and starting to be more sophisticated and nuanced on where you put your processing functions. Look, I don't have time to go into all of these, but there is a number of things to be considering around digital enablement. There's a lot of use cases helping Sort of a holistic calculations of value, whether it be application cost avoidance or integration cost avoidance or digital service, you know, the operational costs and the network cost reductions, all around keeping in mind app performance and operational resiliency and sustainability and security and sovereignty and cost optimization and you know, real time flexible network management, you know, egress optimization, etcetera, etcetera, etcetera. Providing commercial flexibility as well, being able to burst up and burn down. All of these are considerations that you need to be thinking about the planning and super important when you're starting on this journey. And and we are really doubling down, and it's one of our big bets. The last thing I'll leave with you on this is, keeping in mind that we wanna get into the panel fairly soon is If I leave one thing for you to consider around your your your sort of planning to win around here is that there are new models for combined TCO in your planning of how you move, start planning your core and and and moving to cloud. Traditionally, what we've seen is an even in BT and and most of my peers and service providers as well. We've considered the application work stack, the infrastructure layer, and the network layer as discrete towers of value. And each of them have business cases, and traditionally, each of those stove pipes. When they do their business cases, often, they've pushed the pain bubble into the other tower. So, you know, the procurement team in one tower and the finance director for that tower, you know, they're high fiving, thinking they smashed it, And what they've done is actually just moved cost elsewhere within the business. Our thesis with NBT is that Really, customers and ourselves are becoming more holistic. Thinking about it as a a holistic calculation of value And therefore, you need a holistic approach to how you look at your architecture as a single unified field. And that includes the cost that input for each of these areas. So in the application area, you've got the calculations of processor. Disc memory data egress, in the application workload layer, in the compute, you know, in the the legacy completed area, people are trying to be more efficient around power and compute, compute and, cooling that you know, how do you offload to the GPU, the GPU, the Nick, Obviously, the GPU with AI now is becoming super, you know, everybody wants their own AI foundry to do training. So you can have inferred models in everything. And then, of course, in the network space, this network is now holding together. Applications that used to be monolithic that have been exploded and that are all over the place, and you have all these idiosyncratic functions. So the need for a deterministic network that you have full stack of observability is actually really quite important for your operational resilience. That wide area network is performing the function of the data center land of what you thought, which allows you to unlock really nuanced, really valuable decisions around where they are. Now, of course, that last network part people in that area are thinking about latency, egress, transit, circuit cost, port cost, security, and SDN, and all the control overlay controls that you would need be running a modern telco network for own use and also for customers. Now, each of those alone will not win. In fact, each of those are theses is that you probably will end up failing because you've you've missed out the nuances of how all of them interact together. By planning a really well and efficient on your your workload placement and how you connect your network and how you build your network fabrics will have a real impact to your application workload operational costs. And vice versa, if you're not planning in your applications to understand, to be network aware, you're also not necessarily going to get the best outcomes on on, operational experience from those applications. So if I was to leave you one thing, holistic view of planning to win is one of the things that we're building for the future, and you'll be seeing some announcements for us on how we're building our networks and how we're going to serve. Our channels, our partners, other telcos and our customers, And, and and we'll be unveiling that in October as well. So it'll a little bit of a sneak preview on what what are thoughts on this area, which I think is hopefully very topical for the panel conversation that is coming up. So I'll hand over to my colleagues now. And thank you very much for your time. Thanks, Colin. That was a great keynote, and now I'm gonna hand it over to Monica for her panel. This, thank you so much, for the introduction and calling. He did an excellent job making our job much easier because, Colin went through all the reasons. And why why do we wanna move to, a cloud corner. So now what we're gonna do is we're gonna go and talk about the process of getting there. And I have excellent panelists with a lot of experience, in telling us what is that, what is the best way to moved to go through the transition, be ready, plan for success, but also be, realistic about what are the challenges, and how to manage them. Because no matter what you do, change always, brings a challenge. If it doesn't ring a challenge, it's not really, good. And, I mean, it is not really a big change. So anyways, as I said, I have, excellent panelists have four of them. And, it I have, Micah, vice president of engineering and network operations at US cellular. John Isch, in Infrastructure Solutions at Orange Business. Delip Krishna, a director telco cloud solutions and delivery at Rackton Symphony. And Justin Ryburn, field chief technology officer at, Kentik. So what we're going to do is we are going to, go through, q and a as if you're doing the panel. What we're gonna do, is to we're looking for your questions from the audience. So that's what make a panel good. And we'll we'll take them as we go along. So don't wait until the end. We will have questions at the end as well, but, if you get your questions earlier, you have more chances of us answering them. So I'll encourage you to, ask your questions as we go along. Now, before we got into the the Q and A, I would like to ask our panelists to introduce themselves so that you know what is that, they are doing, what is the role in the transition to the cloud core. Mike, why don't we start with you? Sure. My name is Mike Deanhardt. As you mentioned, I'm the vice president for network at US Cellular, we're a wireless facilities based carrier in the United States. I'm responsible with respect to cloud, for developing our strategy. And our implementation plans as our network migrates to a cloud based score. Perfect. John, how about you? Yeah. Hi. I'm a John ish. I'm with Orange Business as you said before. I'm part of our connectivity business unit, and I have a focus on in structure, which goes from, you know, underlay, overlay, cloud security, those kinds of things. And I spend my days talking to our customers who are large global multinationals on what exactly they're doing in there and how Orange can help them through those transformations. And you have a beautiful background, things are not. Delip, how about you? It'll appear. I think you're on mute. Somebody had to be. I'm sorry. Sorry. Yeah. Sorry about that, sir. Yeah. No problem. Yeah. So, yeah, I'm dilip Krishna. I'm a director of telco solutions with the cloud business unit of rocket and Symphony. We kind of help our telco customers in their, cloud transformation journey through, sim cloud, product portfolio. And I've also had experiences in the past working with the cloud native packet core products and cloud native network automation products. So I'm really excited to be here to discuss this topic on cloud, something, which I'm really passionate about. Excellent. And, Justin, what about you? Yeah. Justin Ryburn, as you said, Monica, field CTO here at Kentik. Kentik is a network observability plot. Form. So, we help a lot of our customers as they're migrating, workloads, specifically these these core source rider workloads into the public cloud have, visibility and observability into what those network traffic patterns look like. So I spend a lot of my days talking to customers hearing what challenges they're having and working with our teams to improve our product to better serve those needs. Excellent. So you guys are all looking at the, you know, you're you're involved in different ways in the transformation to the cloud core. So let's start sort of from the beginning. We heard from from calling what is the transition, to the core for BT, but maybe you can tell us why is that we have to move to the core? And maybe your perspective might be a little bit different because, as a greenfield operator, things are a little bit different. But so, what does that mean for you and what, why do we need to move to the court? Or in your case, you're now moving actually to w to the cloud core. You are that that was the the the way you started. Right? Yeah. So I believe this question is to me. Right? So, so we all know how does the rapid growth of data, video content and all of this is happening in the application vendors have done an excellent job innovating at a rapid pace to meet those requirements. And, this also means that the network has to be really ready to, take on those demand on that particular demand as well. Right? So the from a requirement perspective, you know, what we're looking at is really, a rapid feature roll out. So they're gonna have continuous ton of features. I'm not they're gonna just bring in more and more features, and there's that the network has to hold up to that. That is gonna be a demand for, scaling as, like, we there'll be vertical scaling horizontal scaling to meet the traffic requirements. Then also the jog jog distribution. Right? So where where you may have a lot of edge use cases that may be coming in. Well, that's more from a requirement perspective. But the kind of advantages that we seen not not just within RACoten, but also, from the standpoint of some of the other networks that we are working with, other customers that we are working at. Is that, you know, the cloud is a natural fit for all these diff different requirements. But along with that, the additional advantages that we do, that we see is the the cloud is really, an infragnostic thing. Right? There's a lot of TCO savings that you're gonna see. The cloud is really infragnostic, where the the the customer can go and choose their own hardware. The applications as they're moving more into a microservices architecture, moving away from the legacy three tier architecture. We can see that there is a reduction in the overall data center footprint as well. Right? We will only need to grow as much as you want based on your traffic. So that's another another savings that you're gonna see. And, the another point is the the automation itself. Right? It's one thing that we have seen is the automation is very much synonymous to the cloud. So you're gonna see a lot of operational and the deployment aspects getting automated, and that again is going to, you know, help you, kind of handle a lot of the load that is coming in the future. You're gonna rinse and repeat a lot of your processes. So that's another important aspect that we have seen. So, you know, net net though to summarize the the elasticity, the automation, and the related TCO savings are the the main reasons. I think one should definitely look at moving into the cloud. Absolutely. So there's plenty of reasons to move. And, I would say that in a way, it's kind of an unavoidable move. The question is how quickly do you go there? And, Mike, can you tell us, you know, how much of an urgency is there? You know, do we have to just stop whatever we're doing? Move to the move to the cloud or Is there more of a, sort of phased process in there that you see we can go through? Sure. Speaking as an operator, the experience of customers is the highest priority. And so we do not view it as a burning platform that everything has to stop and move to the cloud immediately. Rather, We are moving to cloud currently, depending on the platform and the service in a very deliberate and measured fashion, with a focus on making sure our customer experience remains robust and reliable. What I will say is, that the the urgency comes from, knowing that, Delip, as you said, cloud allows us to speed up automation to improve monitoring and performance management, to improve how we detect anomalies in the network experience and it changes the cost structure of the business as well. So in a few cases, we have platforms that there are no alternatives that do not rely on some form of cloud enablement, and for those we're moving very quickly. And others, we're testing in small stages to make sure we protect the customer experience. Excellent. And, now, let's go into somewhat of a different, area which is the enterprise. So, for the enterprise, there is, there are also reasons to move to the to the cloud core. How critical is it? Because clearly the enterprise of the the the role of the core in the, the role of the core and the role of the cloud in an enterprise, it's it's different. John, can you tell us a little bit about that? That's your specialty. I think it Well, it depends largely on the customer kind of, what and I and it's not, well, some of it is vertical based. So most of our customers have moved the majority of what they have to the cloud. And, you know, when I talk to customers, that can be I say majority, that can be ninety percent of their load, workloads are in the cloud today. So that's I would consider that sort of a high watermark. I don't think any any organizations in the near term is gonna get to a hundred percent. But, you know, we have a lot of customers who have cloud first initiatives or and, you know, sort of can we make this work in the cloud first? And if not, then somewhere else. However, we do have organizations that I work with financial services is an easy one to, kinda segment out who do not want their, their workloads operating in the cloud. And they often do not operate a cloud based infrastructure to any great degree other than, you know, maybe some CRM or something like that, but their main business applications do not go to the cloud. I think what has changed, dramatically here is that, or maybe What has changed dramatically is where the workloads are. What has not changed as quickly or has not is starting to catch up now is how do I design a network to optimize the functionality that I'm gaining out of the cloud. Right? And, I see we're working with a lot of customers on I guess what I would call SD WAN SD WAN one dot o transformations to SD WAN two dot o transformations and really changing the way their networks operate in order to accommodate that. A big part of that is security. And I know we'll talk about security later on, so I won't say too much now, but There's a lot to, just moving everything from the data center to the cloud is only the first part of the transformation. There's a lot more behind that that has to happen. And I think in some cases, customers are very far ahead of that. In other cases, it's something that they catch up on as as, we go forward. I think we're gonna use or I'll use the word transformation a lot over the next forty five minutes, but a lot of these things are not, you know, flicked flick a switch and everything's done. It's a transformation. It's a process. It's a journey. And if anything that's a journey, you have to understand where are you now? Where do you wanna go? And then map out how you get between those two points. Absolutely. And and because it's a journey and a transformation, there is a lot, as I said, there is a lot of certain aspects that come into it. So it's not just as simple as move to the cloud. There's a lot of implications about it. And, Justin, maybe you can tell us about all those challenges from, you know, being involved into, helping your clients during this transition. What do you see at as the main challenges in the bigger picture? Sure. Yeah. Yeah. I mean, I think that they they the details can vary from organization to organization as I think, John was just alluding to, but If I were to put them into high level buckets, I'd say they fall into cost, performance, and security. Those are the three that are kind of the summary of things. Cost, you know, I think everybody has this bias in their head that when we migrate the cloud, things are gonna be cheaper. And for some organizations, that can be true. I mean, if you are able to shut down a data center that you're currently operating and you had huge cost for operating that data center, whether it's power, cooling the staffing to run that invest continuing investment in building out and, increasing the size of that data center. There can be some cost savings. But I think what what we find when we talk to customers is they're a little surprised once they, you know, move these workloads to the to the public cloud that they don't wind up saving as much they thought they were going to. And there's a lot of reasons for that that we, you know, maybe talk a little bit more about as we go. But, you know, cost tends tends to wind up being a concern for customers as they as they get migrated to the cloud and then wanna figure out how do we how do we optimize the cost going forward. Second one being, performance, again, is I think some of the panels have already alluded. When you when you migrate a an application or a workload to the public cloud, you have to do it in such a way. You have to plan it out in such a way. That all the pieces of that application migrate together. Otherwise, you're gonna have high latency. You're gonna have problems with the connectivity in between them, and that's gonna cause performance problems for those applications. So it has to be well thought out. It has to be well planned so that the applications perform the way that you want them to when they're they're migrated to the public cloud. And then the third aspect there was security. I think, you know, there there are ways to secure app applications running in the public cloud and all the public cloud providers have some best practices that they have published on how to do that, but it's very different than the way it was done when you had these workloads running in your own data center. Right? The the notion of a security perimeter that your firewall occurs for you, doesn't really exist in the public cloud. That's not the way they secure workloads in the public cloud, so you have to make sure that you rearchitect things such that the security posture that you had when those workloads are running the data center is maintained in the public cloud and have some way to validate that. Right? And that's where we see a lot of a lot of our customers looking for solutions is okay. I think I know what my security posture is and how what's being enforced, but I, you know, I need some way to double check so that I'm not exposed. Justin, I just had a question for you. I think that, you know, we're getting to the point where cloud is being done in a deliberate manner. You know, the the entire multi cloud environment is starting to become more deliberate. But I think There was a long time where cloud was kind of, haphazard and done not with a central IT kind of focus of like you'd have in the data center, right, where everything is done in a very specific way. And do you think that's true that we're getting to the point where things are are that that, deterministic way in which you would deploy cloud is be is becoming more the norm rather than that kind of haphazard way of the past. Yeah. Definitely for a long time, there was the the concept of shadow IT, right, where there were departments within the organization who spun up cloud infrastructure using a credit card and, you know, probably didn't follow the proper, IT guidelines that the organization had I do think that now that the organizations are becoming more open and seeing the value of migrating stuff to public cloud that those migrations are a little better thought out. You know, there's probably still some amount of work that has to be done to clean up some of the shadow IT to get a handle on what is out out there. How do we make sure it's following our best practices and the way we wanna architect these things. The other thing to the multi cloud that we see a lot is organizations who don't intentionally have a multi cloud strategy. Right? Like, they may say, hey, we're all in on I'm just gonna pick one, say Google Cloud. Right? And either they acquire another organization who was all in on Microsoft Azure. And so now they have, you know, two clouds. They have to support for some period of time through acquisition. Or, you know, they had a shadow IT where Google was their approved cloud provider, but they had certain of the organization who just found it easier to spin up Azure because Microsoft gave them some free resources with their o three sixty five spend or whatever it may be. Right? So they have both in their environment, whether they intend to or want to or not? We're also seeing customers make by, decisions on software as a service and, infrastructure to service in different directions. But then things get cloudy if you can apologize for the pun. But, you know, things start to get a little bit cloudy in between there and and what is which and, you know, did we, you know, so anyway, I've seen, but I've seen that multi cog go that way too. You know, we're I have different and they start to bleed over into each other. Anyway, for sure. Yep. Yeah. No. And it's it's really it's an interesting, set of considerations and especially in telecoms and wireless, where we are using we're sort of adopting an IT model, which started not within telecom. So we're kind of, in a way late to the game, but that also gives us the opportunity to learn from what other what what happened in other sectors. So now let let me we're gonna go to security because I can see there is a interest from the audience as well on that. But before we do that, maybe Dylan can tell us how to get ready to succeed. You have a tip, like, what is the the most important thing that you can Do in order to make sure that the transition complex as it is is going to take you where you wanna go. Is there Sure. Absolutely. Right. So, so there are several steps that organizations can take in order to have this smooth transition into the cloud. Right? But according to me, the number one in that list would be a precise planning. I think some of the gentlemen here spoke about that too. Right. The the first and the foremost, and that would be the choice of the right cloud platform. Alright. And this largely depends on on the use cases itself. Now, for example, if you're looking at the core data centers for telcos. Right? The requirement would be around horizontal scale. They could be networking related requirements, specifically. Like, we may have to support multiple different CNIs for packet core applications. There could be a demand for effective storage management through cloud native storage solutions. And, also, maybe, for, features such as backup and restore of the applications, specifically for those, in the BSSs and the OSS arena. Right? So there may be a lot of those requirements. Now those could be from a core data center perspective. But as we move to the edge, due to the sheer scale, of the, the edge network, you'll need a platform, which is very lean. Right? And it does not eat into the resources of the applications. Right? Because you have so the your network base is gonna be so big that you may not be able to, you know, allocate a lot of resources to your platform itself. So you need that, the lean nature of the form. Right. It may there may be a scenario where I mean, we, ourselves have seen that there may be scenarios where, just a single note deployment. With the hyperconvergences is gonna be the requirement. Right? So, the the lean is the go at the word basically when it comes to the edge. Now if you see, there are multiple requirements already from a, from a cloud perspective itself. Now the second point that we have to look at is how are you gonna use automation? Right. Now already, the cloud, the, like, like, with the topic itself says that you may have actually moved some of your smaller applications into the cloud, how do you now deal with the critical ones? And then we have you may have had some learnings as part of your smaller applications. Right? And, and you may have created some, set of procedures already. You take those learnings, takes the the the extra steps that you've created and then you create a mop out of it. I mean, usually, there's all the work that you've done translates into a mental procedure. So you take that mop out of it and then put it into an automation framework. Now that is where the choice of an automation framework also becomes important because, you cannot because the use case are gonna be very dynamic. You cannot be waiting the product change to happen, which may take from three months, six months, whatever it is. Right? So you may want a platform which is very flexible where you can just go and write your you know, scripts, bring up your automation and just try and, you know, finish off the work. Now this automation itself could be from the bare metal. It could be about app app cycle management could be about building the clusters, and and many other use cases. Right? So, these are some of the things that we have seen at that that we have done ourselves also, which have kind of, where, you know, we have gone ahead and succeeded not only within Rakuten, but also some of the other customers that our company is working with. Right? So, these are some of the steps that I think are very, very important for the organizations to follow to to kind of succeed in their cloud transition money? If before we move on real quick, you said something that I wanna I wanna underline a point that you said because I think it's really important from an operator perspective the maturity of the automation framework has to move in tandem with the cloud with with the migration to cloud. Cloud by itself just adds complexity because nothing else comes out of the network. You just add cloud on top of it. And without my without automation moving at the same pace, which is sometimes just as complex to enable within an organization, the costs are actually higher with cloud because of the people you're throwing at the problem to manage So I you said that. I think it's a it's a really well point a point well made. Exactly. Exactly. Right. No. I was I was, like, I was agreeing with, the the same point, right, that basically, as you learn, because the the cloud is gonna be dynamic. Your use cases are gonna be dynamic. And as you learn more and more from it, you try and close that loop through the automation so that, you know, your the next set of use cases become a lot more easier. Alright? So that's something that we have experienced ourselves. Go ahead. And I think automation, you know, you you that's, it's an easy word to say, like, you know, automation is gonna make, you know, like it, like, it's, I I think there's a question about AI later, and I think there's a very similar, oh, AI gonna solve everything. Automation's gonna solve everything, but there's a lot that goes into automation. Right? It's not push a button and everything happens. You have to develop the automation for yourselves. And from a customer perspective, that means different resources than I probably have on staff today. Right? I'm bringing essentially programming resources in pro potentially. And I have to not only bring in those resources, but the or train my existing resources and programming, which is probably a very hard thing to do. I I was talking to a customer two weeks ago, said that act thing. It's really hard to take a network guy and make him a programmer. You know, it's just not it's a new skill set, you know. So but You know, anyway, it's just a resource thing. I think you can't it's easy to say automation, and that just kinda flows out like it solves every problem. But there's a lot that goes into automation to make it, do what you want it to do, you know? And maybe that sounds simple, but it I think it's, it's a bigger ask than a lot of when I talk to customers just starting out on this journey, not recognizing how difficult it can be to to find those people and keep them. You know? Yeah. And actually, finally, another thing is at the same time, moving to the cloud might give you the opportunity to in introduce automation in a way that you wouldn't wise. So it sort of works both ways because if you have to automate the existing core, it might be kind of a waste. So you just do the whole thing together. So it's a way to push you towards being a little bit more aggressive on the automation side. So so, Mike, do you see that to be like that that pushes you to do automation to do so it is the is the transition to the cloud going to push it to innovate more than you would otherwise? I think both Because if you have to do it, might as well do it well. Yeah. Both both automation and cloud movements in separate swim lanes today. The challenge is they've gotta be linked to one one automation enables cloud and cloud by definition enables faster paths to automate. And so they have to move in tandem. We're working both. A lot of it comes down to social development and training. And and I I'm a strong advocate. And I know we talked for a minute about, organizing the cloud migration very carefully and through a a strong planning framework And I agree with that. I I agree with the centralization of planning for cloud migration, but I would advocate for the decentralization of training and development and cultural engagement and change all the things that underpin migrating to the cloud. So automation is one of that. We we've decentralized our automation training. We don't have one group that does automation. We've we have dozens of people that are, that we've trained in automation embedded within each work group because that's the only way to make it work and make it grow fast enough. Yes. And you mentioned culture which is, an important, topic. I hope we get there later, but let me get to the first question we have from the audience. And I'm reading it to you. It's, my concerns are security and performance for the five g core. The routing into and out of the cloud with that lenses How is that benchmark and five g is all about throughput minimal QA QOS based latency and how is that gonna be impacted? I guess at court. So there are two questions here. So the first one is about security. And, and Justin, maybe you can tell us a little bit about how do we make sure that as we move to the court, we do not compromise security? Yeah. I think, there there are a lot of logging tools. There are a lot of, approaches that the cloud providers give you. To be able to see what your traffic looks like, what, act what's being accessed, all of those type of things. So just like you would in a in an on prem you know, mobile serving center or on prem data center, you've gotta keep an eye on those things and make sure that you don't have breaches that you don't have traffic that shouldn't be application shouldn't be being accessed or traffic that shouldn't be there. They're just a little bit different in in what you're looking at but the types of things that you're trying to find really don't change that much. It's just a matter of making sure you're keeping an eye on those those things and looking for those anomalies. Yeah. John, do you wanna add some is there any specific? I mean, I guess the enterprise might have some additional concerns in terms of security. Yeah. In fact, I think that there's probably two different things that go on with global multinationals when you look at cloud security, there's the optimal way to do cloud security, and then there's the way we've always done security in the past. And those two things have to come together in order to get to where you want. And, you know, look, I'm not I don't think anybody who's trying to sell security to a large multi and walks in and says you're doing security wrong, and here's the way you're gonna do it. Yeah. You know, that's the end of that sales call. Right? That's the that's not the way large global multinationals are gonna make decisions about security. So, I think the the challenge is that a lot of times when we see large global multinationals make decisions about cloud security lags behind, and we end up in scenarios where we do have, you know, whether it's a I saw us on the the BT presentation earlier, a CNF, a carrier, neutral facility, or a multi cloud, security platform or something like that. It still becomes a, a choke point for cloud traffic. Is it optimal No. It's not optimal. Is it the best way that we can, deploy for a customer who has those requirements, then probably. Right? So we have to meet we have to exist somewhere between the two. We can educate about Optum the optimal way to do it, how these things are gonna become efficient. But we have to operate within the customer security platform or security requirements as they exist today. So I always Alright. Alright. Number one, anytime I'm talking to a customer about SD WAN or cloud transformation, I want security in the room. And I, and I think that's a mistake that's made sometimes in those kinds of, can be problems in, in those conversations is securities left out of the room. And then what happens is you know, whatever discussions you had become absolutely mute move down the road when security comes in and says, oh, no, we can't do it that way. So it's really important that security is in the room, but we have to learn to work with them and say, how do we work together to come to the the best solution for everybody in the environment? That's, and, you know, I work I'm working on an RP right now where the customer said, this is what we need for security today, which is largely replicating a data center based security platform that they have in place today, because that's the easy button, right? But they want to make sure that anything we deploy has the capability to do more advanced cloud based security in the future. And you can see in their minds, right, what the RFP is doing is set setting the stage for a transformation and security from where I am today to where I want to be tomorrow. But, not all companies can instantly flip that switch to say, okay, that we're going to go to a new security paradigm. It just doesn't work that way. Yeah. So I guess it's, you know, I guess that's one of the tips for, the transition. Keeps security involved from day one. You know, from the RFP process and not not just at the at the end. And I think we moved in not to interrupt, but I think with, Sassy, the whole idea around Sassy is kind of bringing security into your SD WAN transformation, but it also applies to cloud a hundred percent, you know? Absolutely. And, Mike, maybe you you can comment on the performance part of the question as we go to the five g core. Is that going to limit the kind of performance, that you have in, you know, improvement in performance you have in the in the RAM. Because is is the cloud slowing down everything else, basically? Yeah. Generally speaking, no, in part because it's a directive. Right. We we did the cloud migration cannot impact the customer experience on the rant. And and as as the audience member noted, that's that that experience level is a critical threshold for successful retention of customers. And so the way we manage it very practically speaking is, the most sensitive network elements, the the IMS core, essay core, those types of network elements, we maintain in a cloud environment on premise. And so it's the first little baby step in cloud. They're they're they're certainly cloud enabled. They're in cloud environments, but they're on prem. And so we manage the performance. We manage the latency. We manage we manage the security in our own data centers. All those things we've we've managed by taking the the question mark, of external latency, external performance, external security out of the equation. We will use some applications that go to, either private cloud, off prem, or public cloud. We're doing some of that today, but they're the least sensitive to the customer experience? Absolutely. So no no compromise. Anyway, so, anything else that you would like to comment, Adel, would you comment on this? Sure. So I think everybody's been talking more or less from an application perspective. But from an infrastructure perspective, from a cloud perspective, what we kind of make sure is because the cloud is not just cloud alone. Right? It's not the cloud layer alone. There's gonna be an operating system below it, then there's gonna be the hardware below it. Etcetera. So for us, it's the whole stack that have to be secured. So, I mean, before the applications come on board, on the cloud. So what we kind of try and do is we try to stick to all the CIS benchmarks, right, from the operating system perspective followed by that on the cloud from a Kubernetes perspective. We try and make sure that we are always hardened, on every front. And also there are other platform capabilities right, like the service mesh and all that. That's something that we bring, along so that, you know, we offer all those capabilities to the applications that they do not have any kind of concerns when it comes to security. That's some just just my my two cents from an intro perspective. Absolutely. Now we talked about, automation and how this is crucial as we move to the cloud. How important is AI in driving this automation? And maybe you can comment a little bit more specific about what you know, automation seems like it's a good thing no matter what. Right? But can you be a little bit more specific about what kind of automation? How is that going to help and how is AI going to, drive that. Who wants to take that? I can I can start and oh, go ahead, Justin? Go ahead. Go ahead, Justin. Yeah. Oh, you go. That's it. Awesome. Yeah. I mean, I think Sure. Sure. I, you know, I think AI is definitely an interesting topic. There's a lot of buzz in the industry about it. I think there's definitely some we're we're starting to see where there are some areas where it might be able to be applied now...