Unleashing the power of private 5G, Edge, and Generative AI: Insights from MWC Barcelona 2024

So here we are. We're back. TGI, the great indoors at MWC twenty twenty four this year. And it's Monday afternoon. It's the busiest day, I always say, of the show. And I'm now gonna be joined by a fellow Brit. His name is Chris McKenna. He's the head of business development private five g at AWS. Oh, how great. Welcome. Welcome, Chris. Thank you. Thank you. Yes, it's Monday, crazy show, but super excited to be here. I lead business development for private five gs, which is a big domain here in the World Congress. Lots of people doing it, lots of partners, lots of customers, lots of competitions, it's all good fun. But yeah, the main aim is to really get the word out there that private five gs is big, it can really give customers good, amazing benefits, and hopefully we'll dive into some of those. No, we will for sure, we will. And let's begin now. Give us our listeners a bit of a history of yourself, Chris. Obviously, you live in the United States. Now you're at AWS. But how did you get what was your journey, basically? Yeah. Actually, it's it's interesting that private five gs is kind of like the intersection between telecoms and enterprise IT. I come initially from an enterprise IT background, and I went into telecoms. I did IT inside telecoms for a number of years, probably more than I'd like to admit. And then I moved into telco and telco networks. So I sort of naturally span both, which is why it's kind of relevant that I'm now working in private five gs, because it's very much the intersection of both. So we use telecoms network technology to provide private networks, but we provide private networks for operational use cases inside enterprises. So it's very much that enterprise IT, and that's morphing today to AI applications. See AI and Gen AI spoken about here just constantly in every conversation we have, That really sits across just about every enterprise, industry vertical, and every market segment. So my journey really has been sitting across both of those, most recently specifically in the token network space, in the cloudification of telco networks, and if you look at what we're doing in the public cloud is we're working with our telco customers to take those telco workloads, their own networks basically, and move them into the public cloud. So it's a big change in the industry, I think as cloud plays a more prominent role in five gs, you're seeing more of the network operators than you see here. Just looking around us, I mean, you see some of the logos from Vodafone to Zittimi to Microsoft, all intermingling now, and the convergence of cloud with telco is becoming a real thing. So we have a presence here. We're upstairs as you come into the center, we've got a nice whole bunch of demos there. Three different demo topics. One is about how we're helping to transform telcos. One is about how we're helping them to reimagine their customer experience. But the one that I'm most focused on is how we're digitizing industries. So we're working with the telcos and with systems integrators, with guys like Amdocs, of course, to help their customers to digitize by adopting cellular networks and running artificial intelligence applications at the edge on those networks. Now, I want to talk about cloudification traditionally. We focus predominantly, maybe it's because I'm a ArmDocs in the IT domain. Yeah. Lifting and shifting everything from their mainframes out into the cloud. But the cloudification of the network itself is a whole different beast, right, in many respects. Very good. How mature and advanced is the cloudification of networks versus the traditional IT move to the cloud. Where would you say they are on a scale, you know, as far as parity? Is there any parity or is is is one ahead of the other? No, I think parity is definitely not there yet. Think IT have been virtualizing applications for a couple of decades already now, both in the public and private cloud. I think we started our cloud in two thousand and six, so even our public cloud is already eighteen years old. But network cloudification inside telcos is something that's It has been around. It's been spoken about probably for a good six or seven years now, but only in the last eighteen months to two years are we really seeing this happening. So definitely nowhere near the same level of maturity. But that's changing every day. Think one of the things I've really realized in this show today versus previous years is that it's becoming real. It's not all about proof of concepts anymore. We're no longer grooming the technology. The technology is proven. What we're now showing is real world deployments and real customers really ramping up their use of the public cloud. And we've got some great demos upstairs. We're showing anything from five gs core scalability to IMS workloads to customers leveraging AWS to run their radio networks, which is really interesting because that's true edge. You can't really get further towards the edge than the radio sites. So that's the real difference now, but we can only say that's happening now this year, last year. We announced that DISH Networks were building their whole entire network on AWS a couple of years ago, and that was the first time that's ever been done in the world. So if you look at IT virtualization and cloudification, it's fifteen to twenty years old, and Dish were the first people to build a token And Dish were two years ago. And it wasn't like Dish were moving from the traditional legacy. This was a brand new build based And that again comes into play, so that's what we call greenfield, where everything is brand new. It's an easier environment to say, well, we don't have subscribers to worry about, we don't have a network to keep up, we don't have service penalties if we have problems. We can really build this and architect this and create it from the ground up, so it makes your life a little easier. So really, that's not a very common occurrence where you come across a network, is pure greenfield. The vast majority of the guys here are brownfield, so they do have their business to protect while they're also going through this process. So, it's a little different, and it's definitely a more cautious approach and a slower approach, but we are seeing production networks moving production traffic into AWS today, which is really pleasing, and it gives them benefits. They're able to, all the benefits that you see in IT, so it's more convenient, it's faster to deploy. They'll reduce time to market, reduce cost, you only pay for what you need. You scale up, you scale down when you don't need it. We see interesting use cases where telcos are still deploying things in hardware. They have to dimension that hardware for peak traffic. Otherwise, the peaks come, they won't be able to manage them. So if you were to say to the IT industry now, go and dimension your hardware for big traffic and then leave it there when you don't have big traffic, they would probably laugh at you. Whereas in telco, it's still being done. Our mission is to sort of bring those same benefits that the IT industry has enjoyed and bring them to the telco industry. And then specifically in private networks where I focus, it's about then bringing, through those telcos, bringing those benefits out to the end customers to run their operational workload. Now, you mentioned DISH there, and that's a unique case and a big customer of ours, obviously, in the US. Are there any geographical sorry, not geopolitical, but anywhere in the world? Are there more areas that are embracing this quicker than others? Yeah. And there's also a geopolitical aspect. Yeah. I mean, spectrum is the key here. So we do see a direct correlation between those countries that have opened up spectrum for this purpose and the growth of the industry in those countries versus the countries that haven't. What we see is the US led the way with CBRS and really made it simple for customers to, in an automated way, request and gain access to the spectrum for this purpose. And some of our partners, Amdocs, Amdocs are one of the players in spectrum access management and real other partners like Federated Wireless, for example, who supply that. And it's all done in an automated fashion. So for the end customer, it's great. You can get access to what used to be impossible spectrum to get access to. We're seeing that proliferating across the world now. But there are a lot of countries still where we have to go through CSP partners, the telcos, and we're not seeing the same level of adoption there. I'm not necessarily saying it's because of the telcos, but I think the lack of ease of access to spectrum, it does slow things down. And do you think it will be a wholesale move to the cloudification of the network, or do you think there will be a coexistence with the traditional world? Well, obviously, we would love it to be a wholesale move, but but the reality says that that's that's not the case. I think telcos are heavily regulated. They provide essential services to people. They cannot take risks as some other industries can, which means that we will see a careful migration. I think what we see is telcos adopting use cases that really bring good value to them. So disaster recovery, for example, we've got a lot of customers putting workloads into AWS that are there for disaster recovery purposes. And the reason that's useful to them is that they can deploy it with a minimal footprint and then use our tool sets to scale those when they need them. So they're not paying one plus one. They're not deploying the same amount of infrastructure of deploying just a minimum amount, and then our tool sets can automatically scale those if they have an event where they need it. So there's a really useful and interesting use case that's taking off fast, and that gains trust, then they see how we work and how we can perform in those environments, and then we're seeing more and more customers start to move live traffic in small batches and then ramping up those batches all the time. So no big bang, but that's fine. Risk appetite is just not there to take those big bang approaches. And how is the IPW program bringing benefits, accelerating the monetization of the private networks? Yeah, so we launched this program here actually last year. So we're one year in. And the program was launched entirely based on customer feedback. So customers were saying to us that private networks are too expensive, that they take too long to deploy, they're too complicated, and they involve a lot of technologies that, to be frank, a lot of the enterprises don't understand. So in the telco world, we're all throwing around acronyms because we all know them, and we're all complying to standards that come out from the standards body, from 3GPP. And we take that to the enterprise world, they say, what? So, you know, we're going through this journey of making these solutions more friendly for enterprises. And in talking to them, they've told us a few things. So they've told us that they want them to be easier to consume. Right now, they're too complicated. They want them to be more integrated into their enterprise networks, so the data that comes off them can be used in their enterprise systems. And right now, you you hear people talk about IT and OT convergence. Right now, are two very separate domains. So your operational technology domain generally sits separately from your IT domain. So they really want to bring those together, make them easier to manage, make it easier to see what's going on across the network, both in the IT and the OT domain. And that's really where we started thinking about building a new partner program and creating an ecosystem around this where we can enable all these things. We can make it faster to deploy, faster for customers to understand the use case, the solutions we have in their industry vertical, make it easier for customers to consume it by pre integrating core and radio applications across our edge, by pre integrating over the top applications to process the data. We have a bunch of edge infrastructure solutions that we bring, and we do all that pre integration. And it means that we can work with guys such as yourselves at Amdocs to take all those building blocks, build something fast for a customer, but crucially build something that's repeatable. So when the next customer comes, we say, Well, hey, we've already done this. It's already ready to go, and why not just Drop it. Assume this in an easy way. That was the thinking. Year on, we've got tons of CSP partners, ISV partners onboarded. We've got real life deployments. We've got a great use case down in Mexico with WinAmdocs, where we're doing an augmented reality solution around World Heritage Site down there using web networks. So we're really seeing exciting growth. And just switching to the enterprise domain. Just out of interest, just And and I hear cases, and I was speaking to Dale about the private enterprise network that I put in for a brewery, and they're demonstrating it on their booth and they're using AI with Nvidia and How is the demand from the enterprises to deploy private networks? Because you talked about that it's a bit complex. They don't really understand it. And that's why you went into, you know, the the IPW in many respects. But is the demand there significantly? The demand is there, and we also believe there's a lot of demand that the enterprises don't yet know that they have. And that's part of the purpose of the IPW program is to help enterprises to understand what they might be able to do. So as part of the program, we deployed a portal. And as that portal evolves, it's taking on a more industry verticalized approach where we're exposing these solutions so that customers can go in and see what other customers are doing. So if you are a brewery, you don't necessarily have to be a telco guy to see the benefits of private wireless. We can improve their productivity, we can improve their supply chain, we can deploy AI toolsets, as you mentioned, to have them to automate simple tasks to speed up their factory flow. We can do defect detection at a rate that we couldn't do without AI. We released a case study last year where we deployed a single AWS instance into the edge, and it was analyzing sixty four elements of a production line, that it would have taken the same amount of time for one person to analyze one element. So very simple business case, sixty four times one, or sixty four versus one. So, there's real business benefit today, and I think that's where we've really come a long way in the last two or three years in private wireless. It's gone from this very esoteric sort of imagine what we could do with five gs sort of conversation to, you know, we have real business problems to solve and we have real solutions that are solving those and those solutions are in production today. And and what's the sales process? So just to paraphrase what you were saying, Chris, is many of the enterprise don't know the benefits that this can unlock and unleash for them. And the more you deploy and the more you have these almost standardized models or scenarios you can drop in, the more they'll realize the benefits. Obviously, you play a key part in marketing and and and promoting and making people understand the virtues of of Yeah. Of private networks. What about the CSPs? Are they all in as far as this technology for their enterprise customer base? Yeah, and that's a very pertinent question at the moment, I think. What we're seeing is that many of the CSPs have identified this as a new revenue source for them. Mean, so to talk about Eris Mayer from Verizon, the Chief Product Officer, called it out just about two hours ago. Oh, really? Yeah. And and actually, we have a Verizon demo in our booth Oh, okay. For for the ports industry. Nice. So, yeah, we're doing lots of great work together and and, you know, and and likewise with many of the other CSPs that are here. Because we're working with CSPs all across the globe, I think they identify it as a new source of revenue for them. It's somewhere where they can leverage their network expertise. So CSPs know how to deploy networks, that's unquestioned, that's their business. In many cases, CSPs own spectrum that they can leverage and monetize here, so that's another case. CSPs have had to spend many billions in spectrum auctions to get that spectrum. And they have the relationships. And they have the relationships with the customers, yeah. So the CSP channel is a very important channel for us. We're working with them to build these solutions on AWS using those building blocks that I mentioned before, making it really easy for them to build, and then we go out to market together. And one of the crucial things with the IPW program is that when we build solutions with CSPs, we expose those solutions to our internal sales teams. So we have an industry verticalized sales organization with more than ten thousand people in there who are talking to those end customers every day, and they're gathering that feedback from the customers. What do you want? What do you need? What do you like? What don't you like? Where are your pain points? And then we're trying to solve those with our partners, with our CSP partners and SI partners. But I think one thing to call out here is in every single engagement we have, there is a role for a systems integrator because these these solutions are not easy. So we want to make them easy for the end customer, but they're still complex to build and put together. So we need systems integrators, and that's where Amdocs comes in. So Amdocs can play two roles here, or do play two roles here within the IPW program. Amdocs is actually the first systems integrated to be officially certified and exposed on our IPW program. Oh, wow. Didn't know that. Wow. Yeah. It was about, maybe about six months ago, I think. Yeah. So the two roles are, role number one is Amdocs and AWS go directly together out to customers, and customers buy from Amdocs. That's role number one. And then role number two is helping CSPs to build and operationalize and productize their private wireless solutions. So, you can think of that as a CSP subcontracting the SI role to Amdocs to help them to build and scale. And Amdocs play both today. Cool. So you answered my next question now. I was gonna talk about Amdocs' role, I think that's pretty clear. And it would be remiss of me not to mention generative AI. Of course. That's exploded. Mean, it's seventeen minutes into the conversation. I'm gonna have a sweet step, sweet step and see how long I can keep it to the end. But, obviously, everybody's talking about it. AI is nothing new, but generative AI almost reignited the discussion on AI. So how is it helping you as AWS and the customers that you serve? How are you putting that to practice now? So, and you're right, Gen AI is just everywhere and it's in every use case and every conversation we have with every customer. It really just starts and ends with Gen AI. It's quite incredible. I don't think I've ever seen anything in my career that's just come along and swept everything away. You know what's really interesting about you just saying that last statement? I've had many people who have been on the industry twenty five, thirty years. One question I say is exactly what you said. What's the biggest thing that's come along in the last twenty five, thirty years that's transformed everything. So you've just mentioned, as far as you're concerned, this is the biggest step change, technology, transformational, cataclysmic, call it what you will, innovation that you've witnessed. It is. Definitely. And it's very relevant to private networks as well. So if you think about the purpose of private networks, it's really to deploy something in order for a customer to achieve a certain business benefit. Yeah. Now, know that sounds really broad and vague, but let me give you some examples. If a customer says to us, I'm transforming my organization, I want to digitize, and as part of that, I want to automate more of my factory floor than I do today because I want to yield the financial bottom line benefits of that. Well, how do you do that is the question. And you don't do that just with a network. You do that by deploying sensors, cameras into your production environment, and then doing something with the data that comes off those devices and those cameras. And doing something could mean assessing the images or assessing video and making decisions, and this is the crucial part, making decisions based upon that data. That's all AI. So we're just seeing this massive proliferation of AI through just about every use case you can imagine because without AI, it's hard to show that real business benefit. So what we're trying to do AWS is make sure that everyone has equal access to gen AI and AI technology, because we don't believe it's something that should be in the hands of the few who have the deep pockets and the financial muscle to really exploit it. So our approach is by saying, and it's a little different from some of our competitors who you hear more of in the industry, Our approach is to say to all of our b b to b customers, we will give you the tool sets to to build anything you like, anything you can possibly imagine using AI and Gen AI. So and by tool sets, mean, we have this three layered approach. We build custom silicon for training and for inference. So that's custom chips coming out of our foundries, they go into our cloud, all of our customers have equal access to that now. On top of that, we then have a bunch of large language models that expose to customers through a framework we call Amazon Bedrock, which we released a few months ago. And then on top of that, have a bunch of developer productivity tool sets. So things like Amazon CodeLisperer, Amazon Q. These are tool sets that anybody can leverage to do things really easily. So Amazon CodeWhisperer will generate pages of code for you based upon text inputs. So it really simplifies the process. So what we're seeing is we expose all of that to our developer community and we see a sort of explosion in productivity that we ourselves would not be able to do, that we expose it to developers globally and they come up with weird and wonderful things that we could ever have imagined. So that's kind of our approach. And as part of the IPW programme, we leverage all of the AWS tool sets, so anybody that engages in the programme will have access to all of that right on the game. Cool. Now, this wasn't a question I'd planned to ask anybody today. Okay. But I asked it in the first interview and it gave me such an incredible response and then I got a completely different response from somebody else. So we talk about generative AI, how it's helping the business, how it helps organizations, how it's helping telecoms, how it's improving the customer experience. Yeah. But for you personally, since OpenAI is ChatGPT Yep. What's it done for you as an individual? And and the way I look at this, I used to head up the marketing for big data at Amdocs. Right. And we were always, when it was the top of the hype cycle, searching for the killer use case. Yeah, of course. Right? You remember these days at the top. To me, the killer use case is generative AI and the large language models that are trained on the big data sets that can now be analyzed. But for you as an individual, what are you what have you done with it in the last twelve months? Just interesting or creative or Yeah. Interesting you ask this question because I am doing something this year here, this show, differently from what I've ever done before. Okay. Now, next to me here is a tablet. Oh, well, doesn't work. On that tablet, I have a virtual assistant who's coming to me in every meeting and taking notes and transcribing the meeting and turning those notes into action items at the end and summarizing the content of the meeting for me. I'm not entirely sure how successful I'm gonna be at the end of this week. So it's the first time you've the first time I've done. Okay. And it's the first time that I personally have felt that it's not a toy. It's a real productivity tool that has made a difference to me. So I'm, you know, if I was not using this, I'd have two options. I could be scribbling down notes from every meeting. Or I could wait until the end of the day or the end of the week to do that and forget half of what we talked about. Or you could let someone else do it and completely mess up Yeah. That too. The better off. So, yeah, my hope is that I have, you know, accurately captured everything, that it has allowed me to do that in a super efficient way. It's not replacing me, it's augmenting me, which I think is a key message. Think a lot of people say, oh, AI isn't psychological as well. Sure, there will be some of that, but I think we should think about it as a tool like any other. I don't think Microsoft offers many people's jobs, it probably did a few. But we use it, it just became a tool that became ubiquitous in our everyday lives. So that's where I think Gen AI will be and this is just one example of it. But for me personally, I'm hoping it makes a huge difference to me in the next couple of days. No, but you said something really interesting there. It's not a case of generative AI replacing your job. All you have to worry about is your colleague or the person sat next to you who's using it better than you. That's right. Or who's using it when you're not. And then that being infinitely more productive. Exactly. Exactly. And I think that's really interesting. By the way, I've asked that question eight times already today and every single time I've had a different response. So it's a really cool That's amazing. Yeah. Yeah. We're coming to the end, Chris. I've really enjoyed the conversation. It's gone it's gone really quick, which at this stage of the day is is novel. Now, how we finish the podcast is a quick fire multiple choice called TGI to go. And I'm just gonna give you You probably heard this on Radio One many years ago when you were in the UK. But I'm gonna give you two options and you give to me your preference and it just helps our listeners understand you a little bit and we have a little bit of fun. So if you're ready, Chris. I'm ready. Let's do TGI to go. Who would you go back? Favorite city, London or Paris? London. And how often do go back, by the way? I've only been in the US for a year and a half. Oh, okay. So, and I've been back a few times. I'm actually going on Thursday. Oh, really? Looking forward to that. You know, I went back for the first time in eight years last Wednesday. Oh, wow. I hadn't been back for that long. Blew my mind a little bit. I didn't go to London. I flew into Manchester and then I went to visit my mother in North Wales. But it was really good to be back. Yeah. Well, it's always good to be Question number two. And no one has given me an answer to this. I don't know why. But do you prefer ChatGee PT four or the new Gemini Advanced from Google? So you're asking an AWS guy if I prefer Microsoft or Google? Well, it's a well, yeah. It's a good what would be your preference? I'd probably have to say Bedrock from But is that available? Can I go and download? Yes. Yeah. Really? Yeah. Oh, okay. Cool. It's available today. Oh, I didn't know. Because I I listened to a few podcasts last week from the New York Times where they were getting a little bit excited about Gemini, particularly the pay version, Gemini Advanced. And obviously ChatGPT is just synonymous with consumer, yeah, generative AI. So it's good to know that there are I mean, I know there are more options, but of course you would recommend Bedrock. Oh, of course. Yeah. I mean, look, they're they're all absolutely incredible. The stuff that is happening now. I mean, I was looking at the text to video stuff the other day, it's just mind blowing. How fast we can create real content now, it's amazing. I have used them to, you know, not, I was gonna say write documents, but played with them to write documents. And they're just incredible. Blows my mind still. No, it is. So I was at the Boston Consulting Group, Generative AI Summit in New York back in November. Right. And they had a plethora of speakers from all different parts of media and entertainment and how it's gonna revolutionize video creation, movie production. There was an analyst from b c BGC basically saying the next Avatar will come from two kids in a garage in Illinois. Right? And and then showing some examples of where it is now and it's just like, it is mind blowing. The the only thing that scares me a little bit is this blurring of the truth and reality. At some point, you don't know what's real and what's not, and that's a bit scary. And I do wanna get into politics. I always look around when I say that, but I saw Donald Trump talking on one of the cello in Spain now. Yeah. I would spell it out. Okay. I'm going to in the US. But he I thought, my god. That's a deepfake. Somebody's Yeah. And you never And guess what? It wasn't a deepfake. He was actually saying that. But it you know, it made me even now question the veracity of of the sources that you're seeing, you know. The deep tech probably would have made more sense than that. That was not. And and that's why I questioned whether it was real or not. Answer this question. It's an existential question, and obviously, you're you're British. Who was the better footballer? Messi or Ronaldo? Oh, that's one of the all time great dilemmas. Only one of them has lifted the World Cup, and I think that's probably the pinnacle of anything. And he lifted it by dragging his entire team towards that World Cup. He's just incredible. I also think if you look at the way they've ended their careers, I think one of them has sort of slumped off, and the other one has gone and completely changed US football in an incredible way. So, yeah, I think Messi is just an absolute legend. That's a good answer. But and you are a football supporter, football. Very much so. Liverpool. Oh, you're good company. I'm a big Liverpool supporter. Yeah. Good. It's the one thing I miss in Canada. I was a season ticket holder all my life. Nice. And it's the one thing and I missed the game last night because I was here working. Sadly, me too. Yeah. But when I I saw the VAR decision against Van ****, thought, oh, the writing's on the wall here. Right. And then, yeah, can't complain about a hundred and eighteenth minute winner. So we'll take that. Do you know Gary Walker from AWS? I know Gary very well. He's an ex m doctor. Yeah. And he's a really good friend of mine. He's a big Liverpool supporter. He is. He and I live in the same area. Oh, really? We love to go a lot. Yeah. Cool. He's a great guy. He's a great guy. And my final TGI to go today, Chris, it's a bit nerdy, but Star Wars or Star Trek? Oh, that is a difficult one. As a kid, I was a Trekkie without without doubt. Always a Trekkie. I've always loved Star Wars, of course. I think these days, my kids are so into Star Wars that I haven't been able to get them into Star Trek. So our family is much more Star Wars these days, but Yeah. For me growing up, always a tricky. Awesome. Great. And so that's it. I wanna thank you for joining us today, It's been a brilliant discussion. Really enjoyed talking to you and meeting you. I hope you have a great a great week here and Thank you very much. Can't believe it's still day one, but that's know. Good. I know. We're almost at the end of day one. We're almost there. We're almost We're We're Almost at the end. Alright. Thanks. Thank you very much, Chris. Cheers.