Guy Daniels, TelecomTV (00:00:27):
Well, hello everyone. Welcome to DSP Leaders World Forum 2025. Please take your seats, get your coffees last chance for a cappuccino or espresso, but nothing too long. Please take your seats because we are about to start our two day annual event. We are back for our seventh year. I can't believe it's seven years. This journey from CSP to DSP is certainly taking a mighty long time. So hello to all of you here in Windsor with us and also to our online audience because we are streaming all our sessions live. Every session from this year's event will be streamed live. We'll also make them available to watch on demand later. I've got some housekeeping as usual for those of you in the room. Important message. There is no planned fire drill for today. So if you do hear an alarm and it happened last year, if you do hear the alarm, please get out of the room, follow the staff and marshals out of the hotel to a safe place.

(00:01:37):
If you hear a stray beep or seventies guitar riff, that is because we welcome back the charity pinball tournament. We have a range of pinball machines against the far wall over there courtesy of Neil McRae and his friends. An all donations as always to the National Autistic Society. It's a great cause you've given generously in the past of please take part, give generously again because as I said, it's a really good cause and you have supported us over the years tremendously, so we appreciate that. There is also a trophy for those of you who dunno about this, there's a trophy to be won and it would be nice to have a different champion this year. So please take part. Okay, let's get some slide where up because as always, it would not be possible to status this event without the support of our sponsors. So we do thank our sponsors very much for believing in what we're doing and playing their part, and that goes to all our delegates as well.

(00:02:40):
It's your support that enables us to live stream this to our global audience. Now the program for this year is a little different to previous years. You asked for more sessions, so that's what you're going to get. You also wanted to hear from specific industries. So we've introduced the market perspective sessions and as you can see, we have three sessions for you this morning and further four sessions in the afternoon. So you'll be needing a signature cocktail later and more on that in just a moment. Day two, similar pattern to day one, we have an extended discussion that sets up the day and then we follow that with six further in-depth sessions. Full details are on the telecom TV websites along with speaker information. Now for our online viewers, particularly for our online viewers, we are once again broadcasting our live extra shot programs during the conference breaks.

(00:03:47):
And this year we welcome Tony Pulos as the host of Extra Shot. Tony has a little studio set up in the corner of the room and if you've got any questions, send them in to Tony. He'll be using them for his extra shot programs. There's a q and a form on the website for you to do that, and that is where you will also find the viewer poll. We have two questions for you this year, one per day. And the question we are going to ask today on day one is this, which strategic priority will have the largest impact on telco profitability by 2030? You've got three choices. We've given you three choices, but you can only select one this time, only one choice. Now they all say zero at the moment because we're hiding the data for now. But we'll check in on the progress of the voting during the keynote session that's coming up in a moment and also at the end of the day.

(00:04:47):
And for those of you who are here with us in Windsor, you are invited to our evening reception this year. We are very pleased to announce that it's presented by Togi. A garden affair starts at five 30, so that's pretty much right after our sessions end. And the venue is the orchid tea room, which you'll find in the lobby of the main hotel at the very front. We promise you a jolly good time and there's more because Togi have also invited us to the library libations after party where they have a surprise or two for us. However, before we start DSP, we haven't got a special announcement from us here at telecom TV because we have a new event for you. So please if you can join us in person this October for the AI Native Telco Forum in Dusseldorf in Germany because we're expanding our popular online summit into a two day in-person event. More details on that tomorrow. So contain your excitement, hold it back please, and having cranked up the suspense and the excitement to almost fever pitch, let me introduce my colleague, Ray La Matera, editorial director at telecom tv, who is going to walk through some information with you. Ray,

Ray Le Maistre, TelecomTV (00:06:12):
Thank you. Thank you Guy. Good to see you all here today. So I'm just going to talk for a few minutes to start off today about our DSP leaders reports. So telecom TV began publishing our DSP leaders reports in late 2023 with the idea of providing a regular series of insights into industry trends, often based on surveys of the whole industry or subsets of the industry, or sometimes just the members of the DSP Leaders Council, which was also launched in late 2023. Now during the past 20 months, we've covered a broad range of topics from network APIs to green network strategies to open, ran to digital support systems, even quantum safe networking, plus of course a number of reports looking at the different aspects of AI in telecom. Now because the reports are part of our broader DSP leaders strategy, you'll also see a lot of links between the report topics and those of our digital summits.

(00:07:26):
And of course the sessions we have here at DSP Leaders World Forum. Now we've published almost 20 DSP leaders reports so far, as well as other focus publications such as the trailblazing series of six G reports that guy published this year and which you'll be talking about briefly tomorrow. Now, one thing in common across all of our industry reports is that they are free. Anybody can download them from the telecom TV website. Now, I wanted to give you a quick flavor of the kind of insights that we glean from our surveys and industry analysis Once a year we turn to our DSP Leaders Council members who represent the network operator vendor, large enterprise and analyst communities. I'll ask them a broad range of questions about trends in the telecom sector for our annual industry vision report. I'm going to share a few charts from this year's report to hopefully wet your appetite for some of the upcoming sessions.

(00:08:33):
Let's start with enterprise services, A topic that will crop out regularly throughout this event, not least because of the industry vertical sessions that we've put together this year. So we asked our council members whether service providers should focus their new service development efforts on the B2B sector. And as you can see, the result was very clear with 75% believing r and d for most services should be focused on the needs of enterprises. So please keep that in mind over the next couple of days. Now we don't have a specific 5G session during the course of the next two days, but of course 5G return on investment remains a very hot topic in the industry, especially in CFO offices. So we asked our councilors whether telcos will ultimately achieve a decent return on their 5G investments. And as you can see, the council members are broadly optimistic, but the jury is still wavering.

(00:09:39):
We do though have a six G session tomorrow. So let's have a quick look at our six G question which asked whether this telecom sector is doing enough to prepare for six G. And as you can see at the beginning of this year, our council members were quite divided about six G and it'll be interesting to see if the 35% of counselors who believe that six GR and D is a waste of resources either gain or lose support when we repeat our survey early next year. And finally, I want to turn to a couple of results from our most recently published report, the Cloud Native Telco Market perception report, not least because we have a cloud native telco session later on this morning with our industry perception surveys, we're really asking the network operator community for their impressions rather than hard data. But as we know, perception counts for a lot.

(00:10:41):
So we asked our respondents, all of whom were from the network operator community and most of whom hold influential roles at their respective companies about which of their peers they regard as cloud native telco pioneers. They were able to vote for as many of the suggested list of operators as they deemed appropriate and could also suggest others. And I should note that the results you can see here do not reflect the demographics of the respondents. It wasn't a case of just everybody voting for their own companies. In fact, far from it, and this is just the tops seven in the list, there were many more that were voted for by the survey respondents. Now, we also asked our network operator respondents which vendors they regard as cloud native telco sector leaders, and these are just the top six names and I'm sure you will recognize them all, but there are plenty more who receive support from the network operators. For those of you interested in the full list and the dynamics of the results and how they differ to the feedback we got when we first conducted our cloud native Telco market perception study in late 2023, I encourage you to download the full report from the telecom TV website. As I mentioned, it's completely free for anybody to do so. So that's it for me from now. At this point I'm going to hand back over to Guy to get the day sessions going. Guy back over to you. Thank you, Ray.

Guy Daniels, TelecomTV (00:12:29):
Yep, thank you very much, Ray. Well, we're going to start our first session of the day now, so I would like to invite all our panelists for the first session to come up and join me on stage. So if you can make your way on stage and we'll start the session. Thank you.

(00:13:56):
Okay, well, to kickstart this year's DSP leaders discussions, we've got an extra long keynote session that takes the DSP perspective and the value of network innovation for digital services. And tomorrow we're going to have a session that looks at this from the market perspective. Before we start our in-depth panels today, we want to take a strategic look really at the challenges and the opportunities faced by telcos as they continue their transition into digital services providers. We want to ask what are the main technology areas? How can these be addressed and how can the urgent concerns of DSPs, what can we do to accelerate this process? So let's first of all meet our guests. I'm going to ask them to introduce themselves briefly starting on my left with Hannan.

Hanen Garcia, Red Hat (00:14:48):
Hi, thank you for inviting me, having a pleasure to be here. I'm Hanen Garcia, chief architect at the office of CTO for the telecommunications industry at Red Hat. Excellent.

Chris Meads, VOIS (00:14:59):
Hi there. I'm Chris Meads, I'm the go-to-market and commercial director for VOIS Vodafone's intelligence Services company.

Franz Seiser, Deutsche Telekom (00:15:06):
Good morning, Franz Seiser Deutsche Telecom. Happy to be here just in time. Can take you quite some time to get the word at the airport I learned today. So I'm Mr. Deutsche Telekom, just happy to see us number two up here and actually I'm part of this cloud native journey in DEU telecom. I'm heading a department in Germany where we on the core network side, run all our data services.

Gabriela Styf Sjöman, BT (00:15:35):
Good morning Gabriela Styf Sjöman. Also happy to be here, although I arrived yesterday evening, so I did no excuse if I wouldn't have made it on time. And I'm the managing director of research and network strategy at BT Group.

Manish Singh, Dell Technologies (00:15:49):
Good morning everyone. I'm Manish with Dell Technologies. Absolutely a pleasure to be back here at the DSP Leaders Forum. I'm part of the telco services business, five years in the making at Dell Technologies, putting as much sharper focus on telco and looking forward to the discussion.

Guy Daniels, TelecomTV (00:16:09):
Great, thanks Manish. Thank you everyone. Really appreciate you all coming and joining us on this session. So let's get started because we've got a number of talking points we want to get through and we may come to our audience as well for their feedback later, but let's start with AI and AI native. There's this argument that AI native's got to mean automated profit, it's not experimental costs. So how can operators move from pilots of AI everywhere into a truly AI native network? We covered this last year for the first time. We were looking in detail at what an AI native network means. It's going to be important. We're going to find out later how important it is for the coming years, but which competencies are telcos perhaps struggling with, finding it hardest to build or buy? So let's explore this issue of how we accelerate AI within our networks. Gabriela, can I come for you to start with because I know you were here last year with us and you were talking about these issues. What's your perspective on this?

Gabriela Styf Sjöman, BT (00:17:16):
Well, I think there are multiple areas that we need to address beyond the skills, the traditional things that we talk about. I personally think that whenever we assess a technology like AI and the adoption of it, it's important to understand that, to leverage on it. Sometimes it can be evolutionary versus transformational. And if we want to go to that truly transformational impact, we also need to understand what do we need to do differently in terms of operating models. For example, today we've done a lot. I think not only bt, I think many of us in this industry are doing a lot of experimental but also scaling up smaller efforts. But if we want to go beyond that, what I would say component-based automation and truly leverage AI for that service delivery, intent-based networking and so on, we need to address it holistically and that requires that we reassess the operating models, accountability end-to-end for organizations, redesign of processes, modularity, reuse of AI components that we develop. So software development skills, not only the coding but architectural skills and so on. So I think there's a great potential, but a number of these things need to be addressed and we need to approach it from different lenses.

Guy Daniels, TelecomTV (00:18:36):
That's so important, isn't it? Is that your perspective because you're looking in and you're looking at your clients and your customers, the holistic approach we're talking about and looking at it from the different lenses?

Manish Singh, Dell Technologies (00:18:47):
Yeah, I think maybe I'll just start with the AI native part first and for definitional clarity, what's AI native system or for that matter and ai, native telco, which is really building a system or set of systems grounds up with AI embedded and as a foundational technology, not something that's layered on top of it. So starting from architecture design, how you build, operate, maintain all those aspects built around ai, right? That's just the definitional part of it. Where we are in the telco evolution of course with the 5G networks rolled out, that's done and dusted. Where we are going to see is a lot more of the AI pieces coming on top of it. Two particular areas, I've talked about this in the past. We obviously see immediately customer experience and network operations as two areas is where telcos are getting started. As we look further out towards six G and some of the early work in six G starting, one thing that was very interesting for me to see was from the operators through the vendors and a broader community in three GP when they were just talking about what six G should entail, there was almost a broad consensus that AI native is going to be a pillar.

(00:20:08):
So I think that's going to be key where grounds up be design, whether we are designing the radio systems, whether we are designing that corporation core and whatnot. The last thing I just want to touch on this is the falling right how to go about it. And I'll just use our own example. At Dell Technologies customer zero when we started our journey and Michael Dell, he's been very focused on transforming the company to become really an AI native Dell. When we started our journey there were 800 plus use cases. There's no way we can implement 800 plus use cases and fast forward the journey, what we came down to was what is Dell really differentiated about? Where is our business value? What is it that makes us unique? And it brought us down to four areas. Number one, clearly our product development, number two, sales and marketing work that we do, services and support and of course our secure supply chain.

(00:21:16):
And we picked those four areas and we said how are we going to apply AI onto these? And fast forward now all of those four areas, we have delivered the next best action to our sales and services team. We have actually created rack based chat agent for our sales team again as a great aid and prep for the meetings coding assistant for r and d team and really optimizing our supply chain. So I'll stop it by saying this, you can go in many directions with ai, it's a foundational technology, it's a horizontal technology. The question is what's the business value, what's your unique differentiator, where should you focus? Do a two by two on that and plot on the other side in terms of technology readiness and the answers start to emerge very clearly.

Guy Daniels, TelecomTV (00:22:13):
Yeah, thanks meh. And I think we'll find these throughout these two days. The AI is just going to go through every single conversation we have. Franz, can I bring you into this and your thoughts on this topic?

Franz Seiser, Deutsche Telekom (00:22:22):
Sure. So I mean there is no doubt AI will have fundamental changes on the way we build and run our networks and then we are experimenting on couple of things. One thing is still what you find out rather quickly is AI acts on top of your data and of course we're not talking customer data, we're sending back and forth, we're talking all the operational data of our network. The first thing you have to do, make sure you have all the data available in a proper format and make sure you have a proper data architecture. And that's probably an art we only learned recently that you need to spend quite some thought how you architect your operational data in the network because that enables then all the AI tools systems to work upon. It's this famous, if data is not good then you will not get out anything meaningful. So that's the homework you have to do regardless if you talk now six G or current G or whatever G, it's all about the data architecture you need to have in place firsthand. And then I think the opportunities are limitless, right? And agent AI now being the latest nice password being sent around will enable course fundamental changes and always helping us getting better. But currently the focus for me especially in building all that is to provide the proper foundation for making all that happen.

Guy Daniels, TelecomTV (00:24:00):
Thank you France. And probably this leads nicely onto to Chris because you've probably approached this from a slightly different angle in voice. So can we get your thoughts on the AI now?

Chris Meads, VOIS (00:24:10):
Yeah, I think voice we're laser focused on really how can we really scale implementations to drive real value. I think it's interesting, I think organizations in the last couple of years are really struggling to get proper value at scale from gen ai. Some of the research we've got from Accenture looks shows that I think 13% of enterprises are actually getting significant value from gen ai. And I think what we're seeing is people are spending three times the amount on technology of their gen AI budget rather than people or process linking to Gabriela's operating model point technology sets the stage, but it's the combination of technology with people with process that would really drive the reinvention. We need to get that balanced out. That's our lesson is like how do we spend the same amount on the technology as the people and process because we need to get away from this culture of constant proof of concepts. We really need to get to driving significant value at scale and really focus on Productionizing capability and that's where we're focused at voice.

Guy Daniels, TelecomTV (00:25:11):
Chris, thanks so much indeed and I'm sure we'll be going into that in more detail in our AI focused sessions later. Hanan, let's bring you in as well. What are your perspectives on the AI native?

Hanen Garcia, Red Hat (00:25:22):
Well there has been already a lot of things said about it. I will agree with Manish on the fact that we are on a journey. We are yet to see how AI will transform the telco organization. We are at very beginning now. It is true that we're seeing that AI will be present on all part of the organization. That will be the network, but the customer service and you mentioned that as well, it would be everywhere. So I will say that there is one important thing and that is what you started with about the exploration. Some customers are more advanced on their journey, some customers are not yet into the journey and exploration is the very first stage that allow them to understand if they have the capability and they have the skills to actually go through that journey. And as well understanding if they have everything that they need on the infrastructure to do that.

(00:26:21):
And that's come with everything regarding to, for example, a very basic thing is about the data. Do I have the data that I need to do what I wanted to achieve with it? So it is going forward starting very fast because there is no conversation. I will be honest, I have been traveling for a bit in the past month and there is no conversation that I have anybody with A CSP that doesn't end up in discussing the AI topic. So they are very concerned, they are very into I have to do something, I have to start doing something. Exploration is the very first phase. So we have been helping customers on the journey at Red Hat. We are helping customers on the exploration phase. We have the technology and a platform that allow them to go through that journey and we have been helping customers as well on more advanced customers in the journey on what he was mentioning about the scale because that is the other problem. One I have something that I believe is going to make a change for my organization independently is customer service of the network is how do I scale that? And this is where as well we work with partners like like Dell and Nvidia, Intel A MD to try to bring the technology for them to be able to scale that opportunity, that application that they need. Yeah,

Guy Daniels, TelecomTV (00:27:50):
Thanks Anne. Because the rewards are there, aren't they? And there's a lot of research out there on stats as the AM market. There was looking at a few last week potential, 5% uplift in EBITDA with AI native if you believe that that's a very significant number, but the investment level's huge, 86, 90 billion over the next four years and telco ai, so there's a lot of money here, but we have to scale this and we have to look at the business values, which we'll go into in our future sessions. I want to move on to another key topic that we cover an awful lot, which is cloud native and the realization that cloud native can't just stop at the core, we've got to accelerate cloud native, but how do DSPs industrialize this process? And coming back to the points that a couple of you raised there, the scaling issue, moving from proof of concept, permeating cloud native through the organization because we know what cloud native is and what it is not cloud native is just as much cultural and methodology is technology. So Chris, you talked about the problems of moving AI on from proof of concept. I mean is this the same with Cloudnative?

Chris Meads, VOIS (00:29:04):
Yeah, I think so and definitely around how do you scale that cultural change towards CICD. Within Voice and Vodafone we've created this concept of we want to be a pipeline driven organization and so we put a KPI on ourselves as to how much is done by CICD, how much is done by humans And we measure that. It wasn't so long ago, 6, 7, 8 years ago where within the Vodafone opcos we were deploying let's say four or five times a year, big major releases, pizzas being ordered for overnight war room sessions. We all know the stories nowadays we are doing two and a half deployments a year and that's that cultural shift we hyphened off part of the organization started to build a new way of working, a new culture, a new approach, and with it start to shift the entire organization towards that world. That's our perspective.

Guy Daniels, TelecomTV (00:29:56):
Thanks very much Chris. And yet I find the statistic, I think it was Analysis Mason recently or was it last year? I mentioned that only half of telco has really qualify as being cloud native and there's still a heck of a lot of work to be done across the whole telco market. Any more thoughts on cloud native Gabriel?

Gabriela Styf Sjöman, BT (00:30:14):
Well I think that we can take a lot of learnings from our cloud journey in the AI journey as well. I would say that probably we are not the only ones in BT and personally also, I'm not the only one who thinks that we've not done enough when it comes to cloud native by far we've only scratched the top of the iceberg. But I think it goes back to the comment that has been made earlier, which is the objectives. I think that both for cloud as with ai, it's about unwiring and rewiring how we manage a network and the purpose of IT cloud was not, it started with the technology, it's not the technology, it was the objective of that elasticity and flexibility to be able to deliver all these updates in the network, faster time to market, but we kind of forgot about what we need to redesign in order to do that. And you add that to the immaturity of technology from different vendors. I think we've all seen different cloud nativeness from multiple vendors that has added to the complexity. So I think still a long journey to go when it comes to cloud native in my view. But the potential is tremendous and I think we need to accelerate it because if it's something our industry needs is faster time to market, more exploration in terms of the products and services we offer to the market. So it's badly needed

Guy Daniels, TelecomTV (00:31:42):
And minute You were nodding there.

Manish Singh, Dell Technologies (00:31:43):
Yeah, I think lemme share a few things, right? I mean we started on AI data, which is obviously where the industry is headed too, but when we think about cloud native, I would argue that that's a foundational element when we start to think about the networks and the reason is to me is very simple. You can generate all the intelligence from ai, but that is going to be very limited in terms of actionable intelligence. If your underlying infrastructure, if your underlying network is not agile, is not cloud native right Now the question then is, as you said in your question, I mean we've been doing some of it in the core, I would argue there is still a lot of work even within the core but come out of the core and the progress has been slow. We've talked about open ran, we ran cloud, ran edge there, I can literally count on my one hand on my fingertips the number of operators who actually are doing things with cloud, ran open, ran and at some reasonable scale and those two are still very limited, barring up a couple of them.

(00:33:08):
The third thing I would say is in Ray's opening survey that he kind of showed what was very interesting was, and it kind of really brings the challenge out, cons, speakers by absence were some of the biggest vendors in the telecom industry. And so when it comes to network functions, the technology providers, where is the cloud native software stack? And that in the part of the industry needs to catch up. I think they've been too slow to come to deliver cloud native solutions. The infrastructure is there, the tooling is there, it's the network functions, it's the operational aspects that are still to be done. So I think there is a whole lot of work that needs to get done. Core in certain cases is done. Those operators who are now migrating towards standalone, I think that just creates a perfect opportunity for them to really get their cloud native plumbing done and it just cannot remain to core. If it just remains in the core and you're doing siloed implementations, I would then argue don't do cloud native because the benefits of going from a vertical to a horizontal architecture is that agility, that flexibility that we've talked about the economies of cloud, but that only comes when you have a holistic transformation in place.

Guy Daniels, TelecomTV (00:34:35):
Thanks. And yet we know we have to be cloud native for the next decade or more. We absolutely must be cloud native somehow anyone will pick up on this. The comment about we're perhaps slow moving out of the core front. Do you want to pick up on this? I

Franz Seiser, Deutsche Telekom (00:34:51):
Was just so I can only underline all everything you said. I mean starting at the core is relatively obvious because we are talking software everything and I believe everything that is software in the tele net over time will go cloud pipe, CSCT kit, you name it. Unfortunately everything is software, right? You still have some physics to deal with and it's a bit difficult to put physics in the kit pipe. The challenge is now how much can you harmonize the physical infrastructure So you can play on top as just as you said in a horizontal telco cloud that everything that runs on top of this, somehow abstracted hardware can be managed in this agile manner. And as you just said, this is an absolute must have. We can stop immediately talking AI and whatever else if we don't have this baseline in place, that's what for example we are working on as we speak, right? It has to be horizontal, it has to be the same platform for everything because otherwise you have maybe one hardware button, you have one software cell next to the other, which means you have moved maybe a fraction of the way you want to go. So if you want to move to this highly automated AI enabled or even AI native world, the basics is to be cloud native in the first place and that across everything, which is software in your network. Otherwise I struggle to see how you can ever arrive there.

Guy Daniels, TelecomTV (00:36:27):
Yeah, thanks Fran. Hannan, I mean do you want to pick up on the speed because we've got, there's plenty of cloud native solutions out there and approaches and yet the progress we're hearing is still slow, it's not fast enough. How can we push this it

Hanen Garcia, Red Hat (00:36:42):
Being a journey as well? I think I was mentioning that it is take time to introduce technology in the network and we have seen that we are at the fifth generation now waiting for the sixth generation to come and yes indeed because it has been hard for everybody in the ecosystem to get to where people was expecting to be five years ago. Now most of the discussion happened around the infrastructure. When we're talking about cognitive, we are certainly talking about the infrastructure and at Red Hat we have been always believe on this hybrid open cloud where the operators will be able to basically leverage the infrastructure they have on-prem and the infrastructure they have off-prem of the edge and public cloud as well. And that's included. And we have been providing those capabilities, those platforms to many operators around the world that are today in production.

(00:37:48):
That being say as soon as, and I will agree with the panel on there is not yet a definition of what a quantitative telco look like. What there is hope. As soon as we understand that key, the key attribute for those are the agility that we mentioned and that's something that we want out of it. The flexibility being able not only to handle the latest CNF, but as well still keeping up with our existing BNF and even bare metal that's still there and the scalability being able to actually scale as soon as possible and to scale up and scale down as well. So those attributes are there definitely part of what a cloudnative telco is going to look like now I think we have a big opportunity ahead of, that's what I say we have hope here is that the six generation of networks is coming up and we are just starting looking into the use cases. And this is I believe is opportunity for the industry to come together and say, okay, this is what actually Cloudnative telco look like by specification and not just by whatever somebody is doing because all of us are doing something different. That's for sure.

Guy Daniels, TelecomTV (00:39:04):
Yeah, absolutely. Thanks he and Gabrielle, you want to come in on this

Gabriela Styf Sjöman, BT (00:39:09):
One? Just continue. And I think it's about that horizontal. I think sometimes we need to zoom out and say what's the business objective? And we talk about asset utilization for example, which is that horizontal, horizontal telco cloud should help us increase asset utilization, which is horrific in the telco world. I mean it's horrific probably with the worst industry ever when it comes to asset utilization. But to do that you cannot introduce cloud from bottoms up. It needs to be top down. And something that I, I've been in telco, I've been in product companies, probably unit recognizes more where there's a unit that runs enterprise architecture very often in telcos that function is actually missing. So there's really nobody or there are scattered organizations looking at really this redesign of processes because you would need to ask yourself what needs to be true for us to go from 35 40% on average asset utilization of the full network to 60%, which is that the purpose should be hyperscaler economics. And then you say what needs to be true? And it's not only the technology, Chris talked about people and processes once again. So how do you bridge that and then based on people, processes, technology, what needs to be true for that to happen. But I think we often tend to miss out the actual business objective and design around how I meet those business objectives.

Guy Daniels, TelecomTV (00:40:38):
That's great. And this issue on asset utilization we have to explore in more depth later on and we've got a panel coming up very soon on cloud native, so I don't want to steal all their thunder. So we'll leave a little bit of this discussion until then. So let's move on because there's a few other areas that permeate through these two days and through our industry. Another one is network APIs. Obviously a lot of interest in that at the moment, but I just want to briefly touch on this now and Franz, I wonder if I could maybe come and ask you about this one, do developers care about telco APIs? Is there a way that telcos can create some kind of business architecture that will get a share of the API potential? Because it's a question we're going to explore in way more detail in the next two days, but fundamentally, do we have any problems as telcos? Are there any initial obstacles that we need to sort out before we can fully utilize

Franz Seiser, Deutsche Telekom (00:41:40):
These obstacles have been in place? For sure. I think that's the big change to what we see happening now with the more aligned going forward. Are we where we want to be? No, but my impression at least we understood the problem that we have been much too fragmented in the past. It was much too difficult. APIs too complex or to my feeling we are trying to address that now with this common initiative, but the jury is still out. So as you said, again another journey, which we are probably still early days, hopefully we see that picking up. So at least we have couple of APIs out there more to come, but it needs to come on a playing field level. So putting out hundreds of APIs nobody cares about probably is not so useful. So it's very much figuring out what are useful, still easy enough to consume APIs that start to make a difference and people start to see value and start to dig further because immediately you get demand say why don't you have this? Why don't you have that versus you put out stuff and hope somebody likes it.

Guy Daniels, TelecomTV (00:43:06):
Yeah, thanks Franz. And as I say, we have two sessions on APIs this year, so we will go into a lot more detail and in fact it's one of our answer choices in today's poll if you are voting in our poll. So we'll take a look at that in a moment. Well let's move on because there's a few more areas I'd like to cover and another area that we are going to be covering this week is the RAN radio access network. We've been talking about open ran for a long, long time, but I'm just wondering if interest in open ran has waned somewhat recently, are there any obstacles that are still in place that need resolving or should we as DSP, should we be shifting our focus to a more hybrid approach? Perhaps we'll be thinking more about cloud ran, virtual ran and also Manish, if I can bring you in at this point, AI ran because this is something that you and I talked about a few days ago.

Manish Singh, Dell Technologies (00:44:02):
Yeah, my perspective I think first let's start on the open ran. I wouldn't say the interest has dwindled away. There are still a set of operators who are continuing to drive that there have been variations in approach. For example, we all know some of the operators have started with a single vendor to start cloudification of the RAN and an open fronthaul. But the intent is then to bring other radios for the downstream and we will see how that plays out because if it remains a single vendor solution, then why even bother about open ran? So that really is not the outcome or the desired outcome. So that's there. We definitely see more and more move towards leveraging merchant silicon cloud native infrastructure, try to bring the RAN workloads. But I'll tell you this, we were talking just earlier on the cloud native side as well.

(00:45:10):
Same applies to open ran to cloud ran, which is it's a architectural transformation, right? You're growing from vertical to horizontal. And what we've found is within the telco operators, the technology office and the architecture groups, they get excited, they understand the value. Just as said earlier, hey we should be getting business outcomes of better asset utilization and more they get it. But when you come into the network operations, when you come into the procurement teams, they still want to operate the old way. When you are disaggregating the stack, you are disaggregating the infrastructure from the workloads. Your procurement processes need to evolve, your network operations processes need to evolve your organization structures, people's roles, skills, responsibilities, all need to transform. So it's a holistic transformation and that definitely that work is really cut out and has been slower is what I would say. So there's a lot more work I think that needs to be done.

(00:46:20):
Lastly, let me just say, I mean just you touched on AI ran I think early days. We will see how this builds out. I mean there is no doubt in my mind that AI as a technology brings lot on the table for radio access network, you could improve spectral efficiency, you could do better beam forming, you could do better scheduling, you could do better energy efficiency solutions. Very obvious scenarios where there are good reasons for us to be leveraging AI into the radio access network. I'll go back to again the initial proposals in 3G PPP that came up as part of six G. Again, a lot of that consensus building around AI ran that said there are challenges, and I'll just put three big challenges that everybody needs to get their head around. Number one TCO. We've seen that challenge in the world of open ran cloud ran that gets bigger in the world of AI ran number two form factor ran size distributed ran space is limited, power cooling is limited. And so how do you solve all those problems of space, power, cooling, TCO and get to a point where it becomes truly viable at scale. So those are all the challenges and we'll see how we progress on this one. Thank you very much Manish,

Guy Daniels, TelecomTV (00:47:45):
Gabriel?

Gabriela Styf Sjöman, BT (00:47:46):
Yeah, I agree with Manishh, but I also want to say that even when it comes to overrun, I zoom out once again, I think we lost a bit sight of the business objectives. I think it's very simple in general. Either you make money, you save money, there's nothing in between, right? So any technology you introduce has to either make you money through new products and services for your customers. So that innovation or it's really accelerating efficiency and I think the challenge with Overrun has became all of a sudden just about technology on the stack per se. And if you could have said, well I want to save money by just aggregating it, I'll bring in a multi-vendor stack. But that has not happened because desegregation today as it is, it's very immature still and it's really bringing such complexity to the operators that the benefits, potential benefits of desegregating, the supply chain does not compensate for the complexity it brings.

(00:48:47):
And then if I look at new products and services, well at least in bt, we don't see that O ran as it is today brings any benefit to our customers in terms of new products and services. However, AR ran potentially I think we're quite interested could bring some potential new benefits in terms of business models. But I think I always say let's not forget about what are all those business objectives and then what needs to be true. But I think Oran, the way it's going today or has gone is neither making money or saving money for the telcos.

Guy Daniels, TelecomTV (00:49:22):
Okay, right. Well we will pick that up tomorrow morning when we have a dedicated run session. We made a note of that. So we'll see what other guests have to say. Any of my thoughts about the run situation and come in

Hanen Garcia, Red Hat (00:49:39):
To chime in on that one. And one thing that we need to understand is that innovation have a time to market

(00:49:48):
And the fact that being the mobile network has been instructed by the decay rule and being in the middle or arriving in the middle of the 10 years period, it hasn't been as successful or fruitful as many would expect. Many operators out there understand what the benefits are of ran. We have been participating or ran and our goal has been always making sure that we respect the cloudness when we're going to be building what orran is and the specifications of the interfaces to make sure that that will survive the introduction into a cloud native infrastructure. And we're doing the same on the A RAN alliance. That is our goal. We have been working closely with SoftBank into what the early conec look like for that infrastructure. Now I'm going to repeat myself here, but this is where the opportunity is. We have a huge opportunity in front of us. A specification for CG are just about to start. So let's see what we can take out of all this technology that has been built on, ran on ai, ran, and see how we can introduce that into a model that will be interoperable for everybody respecting this cloud nativeness as well on that infrastructure that will be there for the years to come.

Guy Daniels, TelecomTV (00:51:15):
Components we can use going forward from various different areas and build something new for the next 10 years. I'm sure there's a lot going on in research labs and facilities that we may see commercially, but that's a good few years away. And the other issue I thought that came out of what we're talking about here is the need to change systems and procurement and people and processes. It's not just replacing technology, it's a whole mindset. Any more thoughts on Fran want to come in?

Franz Seiser, Deutsche Telekom (00:51:50):
Well I don't think that interest has waned. I mean AI ran probably is a bit overshadowing because this overall AI thing is simply overshadowing all the rest. I think there's still profound work going on. Also in Oran to my knowledge, there just has been a workshop even between three GP and Oran Alliance to figure out a little bit how the split work, which tells me the orran alliance itself is accepted, the value is seen, work is happening, maybe not all things did devolve at the speed as we assume, and especially multi-vendor turns out to be a more challenging one couple of hopes that we've seen some years ago maybe companies digging deeper figured out it's not working exactly that way so we need to adapt. But overall, I think the overall principle of open run is still there, accepted and required. And again, there is a piece of software and if you want to be fully agile and be able to react in due time, also this piece of software, the base station needs at some point in time go into your cheat type pipe deployment operations. Otherwise you have a nice fully cloud native core and then you have a nice black box and tens of thousands of black boxes around it. And then you will still have a nice challenge to run that somehow coherently or you'll keep two completely different islands alive, which I'm not so sure is the way to go.

Guy Daniels, TelecomTV (00:53:25):
I haven't got enough challenges to deal with at the moment. Thank you very much. Without even more challenges, any more comments about the rain you pick up Gabriel?

Gabriela Styf Sjöman, BT (00:53:32):
No, I don't know maybe what we needed, I dunno what you think. Maybe like blueprints because I agree it's not like I say it's there's no valley, I think it's in its current state, it's really challenging. But also these many, many variants. We have lots of engineers who love to engineer their own. That's true.

(00:53:52):
And I think that is a challenge and maybe we need just fewer blueprints. Standardization needs to step in to at least make sure that the operational part of this doesn't become as challenging as it is today. Which is probably also why many, not all but many telcos today who have deployed or ran have done it in a single, it's a single vendor or a near single vendor because it is still very, very complex. So maybe blueprints in standardization is something that there are some efforts around alliance tip and so on, but I think much more to do in that space.

Guy Daniels, TelecomTV (00:54:27):
Thanks Gabriel. That's great. I do want to move on to at least one more area before we go to our audience and that's around zero touch operations. And Chris, I want to come to you if I may on this one. What is the realistic path towards zero touch operations? Where should telcos aim for in terms of fully closed loop autonomy or involving humans in the mix? What are you seeing? What do you see as the optimal way here?

Chris Meads, VOIS (00:54:55):
Yeah, I think for us zero-touch operations for us it's a mantra, it's a way of working. It's a way of thinking about things. Do I honestly think we'll ever get there? A hundred percent, probably not. But it's the culture we're trying to build. So we're really utilizing AI and gen AI technology to just take away all of the manual effort we possibly can. We're looking at it from a, how can we proactively identify issues before they come up? How can we then proactively fix things before they come up, right? Everything we can do, we will do, but the real benefit is then we can get our humans and our really complex engineers working on the complex issues much, much quicker. And if we can get that happening and that really focus, that's when we get good results. So in voice, we've seen through doing this over the last couple of years in this real mantra, we're seeing I think a 90% reduction in MTTR. We're seeing a 70% reduction in incidents and big part of that is the ai, but it's also the fact we can focus our best engineers on the real complex problems. And that's the approach we've got.

Guy Daniels, TelecomTV (00:55:54):
Great. Thanks much Chris. More thoughts on zero touch? We're seeing a lot of AI investment appears to be linked to zero touch operations and there's a bit of a linkage there, Manish.

Manish Singh, Dell Technologies (00:56:04):
Oh, absolutely. I mean you touched on closed loop autonomy and the path towards there. As I said earlier, when especially in the current state of 5G, there's work that will happen inside the network functions, but that I see is more of a six G area. Well, when it comes to network operations and customer experience, those are absolutely the two areas right in front of the industry to really take advantage of ai, generative AI in terms of zero touch in terms of network automation, that is absolutely right here. There are challenges that we should think about of how we're going to solve and not do a pock. One of the things I'm very worried about and as I shared within our own experience also, how do you break out of these POC prisons, right? I mean proof of concept after our concept and nothing really is getting to scale.

(00:57:00):
And when it comes to network operations, zero touch, there is a clear business impact to be had with the technology. The problem areas to think about is start with the data layer. You have to first of all get your data architecture right because without that you're going to be very limited in making the progress. And then of course, I mean leveraging, it's obvious to say leverage all the capabilities now with generative ai, very much here in play as far as driving that automation as far as human in the loop or not. As part of that question, I think my view is it's a journey. Network is the asset for any telco. So I think it starts with human in the loop on the road to that journey. And then eventually as you see the technology mature, the use cases mature, you build a confidence. Maybe at that point in time you get towards full autonomy. Thanks very much.

Guy Daniels, TelecomTV (00:58:07):
Manishh, any more points on the zero touch? Otherwise we'll move on to your audience. Nope, you're okay. Well, we've got about 10 minutes left, so I do want to see if there's any questions from our audience. If anyone's got a question, we have Tony with a microphone who's going to be doing the rounds and we'll be highly selective about who he chooses. Anyone got a question? I'm just going to shield my eyes from the light and I think Tony can see better than I can actually. Tony. Oh, there he is.

Robert Curran, Appledore Research (00:58:35):
Hi, Robert Coren from Appledore Research. Excellent panel. To kick the event off with something that we're noticing in our research is the degree to which telcos are becoming a bit more self-reliant. Again, we've got two great examples on the panel today. Three great examples today, sorry, of telcos. I think looking more to in-house skills versus perhaps what we saw a decade or longer ago, which was where there was a lot of leaning on the traditional network equipment providers to do that. Would you agree that that's a conscious decision by telcos to try to build a knowledge themselves, build it in-house, get hands-on with models and development tools on all the latest things? And does that then tie back to the point that Manishh, I think made about the representation of who are the telco cloud pioneers where the big names weren't particularly on that list. Has that been a conscious effort? Do you think that's a real trend or do you think that's a temporary thing that we're seeing?

Guy Daniels, TelecomTV (00:59:32):
Thanks very much. Great question, Robert. Thank you. Anybody want to pick up Robert's question, Gabrielle?

Gabriela Styf Sjöman, BT (00:59:37):
Well, my view is that if telcos want to shift away from a long trend where the network and especially the operations of the network was not a differentiator to today, the network becoming really enabler for product innovation and perhaps even the product itself, then a key differentiator is going to be how you operate the network, how you deliver those services. So that trend of insourcing and having the skills and build those tools yourself has to be central to that strategy.

Guy Daniels, TelecomTV (01:00:15):
Thanks, Gabrielle. Any more comments for Robert about this? The move towards a soap reliance or in-house skills from Turco? Yeah,

Franz Seiser, Deutsche Telekom (01:00:23):
So what you clearly need to build in-house is integration skills, right? I mean we always talk DevOps. In reality, it's maybe a little bit more in ops as DevOps because we are not developing the CNF software, but we're integrating it, we bringing it together and we operate it, right? So we become our own integrator. Otherwise this wouldn't work. Or you say, dear supplier, you have my whole network, please do it. But then you are back in this dependency where you probably don't want to be. And that's maybe also one of the hurdles why this still takes, because you need to build up the skills. You need to understand what exactly is it your team needs to do, what can you expect to get from a supplier, but this skill to integrate and to run that. And integration also means deliver updates, deliver patches probably on a daily basis. Of course fully automated could otherwise you would be killed. That's a key skill you need to build if you want to run properly in the network. And that's this skill and Mark because this is a very different way of building and running your network versus what has been done in the past. So this is exactly what, if you don't do that, you will be in tough waters.

Chris Meads, VOIS (01:01:41):
I think it's a definite theme, right? I mean the reason Voice created it, right? Not in the network space, but that theme of over the last 10 years, telco is to build up this amazing set of skills and capabilities. Running cross op goes. The reason voice is created and we've started to go commercialize and go external is because we've got all these skills. Actually there's a real opportunity here to commercialize and offer this out into the market and start to share these services globally. I think it's the right theme. I think it applies to network as it does to services in it.

Guy Daniels, TelecomTV (01:02:11):
Great. Thanks Chris. Thanks friends. Great. So I think we may have another question, Tony, do we have another one, Tony?

Geoff Hollingworth, Rakuten Symphony (01:02:18):
Yep. Geoff Hollingworth, Rakuten Symphony. If I ask each of the panel members to imagine just for five minutes that they were either Harry Potter or Khomeini Granger and you had one spell you could cast to change one thing in the telecom industry, how would you all answer that?

Guy Daniels, TelecomTV (01:02:39):
Oh, thanks Jeff. That's a good Jeff question there. We expect questions like that from Jeff really penetrating and it's got its thinking now. Crikey, wave that magic wand. What's the one thing we're going to change? Which is a real tough one. Anyone wants to go first here? Chris? I

Chris Meads, VOIS (01:02:55):
Think I was wondering where that was going for a minute, but I guess look for us, and the reason behind us launching voice last year was the irony of shared services in Telco is we've never really shared anything and it's very insular. And yet that core way of operating a telco, those core services to run a telco, those core capabilities that every telco need, why aren't we sharing those even into the tower codes? Why is there not more sharing going on across the industry? I think that's, well selfishly, that's why we've created voice, but we see a massive opportunity in there in the telco industry. So I think driving more of that sharing approach to shared services would be my spell I guess, or the answer.

Guy Daniels, TelecomTV (01:03:37):
Thanks Chris. That's great going first on that one. Appreciate that. Any more? Spell waving going on Gabriela,

Gabriela Styf Sjöman, BT (01:03:41):
I would like to focus on demand. I think if there's one thing that I would love to see in our industry is that we play a much stronger role in the demand for the products and services that we offer. The reason we're in this price wars of bundles is because it's become a commodity. So the biggest problem to solve for is demand for the new products and services enabled by all these technologies that we're talking about.

Guy Daniels, TelecomTV (01:04:12):
Fantastic. Anymore Harry Potter wannabes on stage that's going to come up with a wish and then I know you're aching to tell us your special wish. I

Hanen Garcia, Red Hat (01:04:21):
Might have many is one thing, and this is come after the discussion we have on the APIs and this is a personal belief. I refuse to believe that the next generation of services will be built on best for internet. So whatever needs to be done, need to be done, that I will expect much more from the service providers on GaN.

Guy Daniels, TelecomTV (01:04:51):
That's a good one. Very good. Thank you very much indeed. Anymore or we're going to call it there? Anyone's going to be brave enough to say, come on Fran, here, there's something you really want to see change in

Franz Seiser, Deutsche Telekom (01:05:00):
This business. I would aim in a slightly different direction if I could. I would look for the metric spell to get rid of the tens of thousands of legacy stuff. We still need support overnight. That's a good one. This holds us back tremendously.

Guy Daniels, TelecomTV (01:05:21):
Great one. Thank you very much Manish. It's going to be Ray to write more reports. Is that your wish? No, sorry Ray. I'll say it. My wish would be for Ray to write more reports because they're so great. So Ray, more reports please. Thanks very much everyone. Look, I see that we're almost out of time. So what I wanted to do, I did promise to look at the poll and see how we're getting on with our poll. So I think what we'll do next before we end this session is go and have a look at our audience poll. Let's have a look here, remind you of the question we've been asking, which strategic priority will have the largest impact on telco profitability by 2030? And we ask you to select an answer from our three choices. Here you are three choices and the real time results are going to appear as if by Magic Potter style. Look at this, here we go. 45% becoming an AI native zero touch operator monetizing network APIs in seconds, surprisingly with 31% and achieving full cloud native agility from core to edge 24%. Okay, well it's early days on our poll. Thank you very much everyone for voting. The polls will stay open all week. Any initial thoughts?

Manish Singh, Dell Technologies (01:06:41):
Yeah, I think it makes sense to me because when I look at the question, the word there is profitability. If it was revenue growth I think may be different. But I think profitability and I think the path with profitability, the poll reflects that. I think it's obvious, but there's a whole lot of work to be done here. So

Guy Daniels, TelecomTV (01:07:02):
That's why we Great. Any more comments that we're saying?

Gabriela Styf Sjöman, BT (01:07:06):
But I agree, but I think also Zero Touch is a key neighbor for monetization of the future. We're talking about all these things, intent driven, network slicing without all of those things. And I also think by the way that we're going to go beyond speed will matter less and less and it's going to be about dedicated performance. And to deliver that dedicated performance quality of service, you need to have zero touch operations. So it's really central for profitability, but future growth as well.

Manish Singh, Dell Technologies (01:07:38):
I agree with that. I would rather even add, I mean for that matter, cloud native, it becomes a absolute e foundational pillar, but that's an enabler with AI native. That's what I meant by the base,

Gabriela Styf Sjöman, BT (01:07:51):
The

Manish Singh, Dell Technologies (01:07:51):
Eventual path to really taking advantage of all the productivity gain that's to be had and improve the profitability.

Guy Daniels, TelecomTV (01:07:59):
Exactly. And the question you're asking that as you rightly identified, is profitability. What's going to have the most impact on profitability? Great. Well thank you much

Gabriela Styf Sjöman, BT (01:08:07):
In a nutshell. We agree with the audience.

Guy Daniels, TelecomTV (01:08:10):
It's good. Good widow end. What to end? Yeah, well we are out of time. It is now 10 30 UK time. So we are out of time for our opening session. Hopefully we've given you a perspective on what's coming up over the next two days and the key issues that we're going to be addressing in more detail for our online audience. Please stay tuned because the extra shot with Tony Pulos program will start in about 10 minutes. So the discussion will continue with Tony. If you've got any questions for Tony and his guests, please do send them in. And as I keep saying, the polls on the website, if you haven't taken part, please do so. It only takes a few seconds. And for our in-room audience, don't forget the charity pinball tournament will continue on the side of the room. The coffee cart is going to open up again and brew up some more coffee. We'll be back here on the main stage in about 30 minutes time. So let's have a round of applause for our guests. Thank you all very much indeed.