How can telcos turn private 5G, edge and AI into new B2B revenues? Orange Spain's blueprint

Tilly Gilbert0:02
Welcome to today's webinar with STL partners. We are going to be talking about the convergence of specialized connectivity, private networks with on-premise edge computing and thinking about how those two solutions are well-served to, are well-fit to serve some of the AI workloads that we're seeing enterprises adopting at speed across Europe today and I'm delighted that we'll also be able to spotlight and discuss a little bit with Orange Spain about some of their own experiences and delivering some of these technologies into their enterprise customer base and we will also be joined by panelists from Dell Technologies and from Nearby Computing as well.

Tilly Gilbert0:53
Just in terms of housekeeping very quickly. I think the main two questions that we get asked by those that join and listen to our webinar series are will I be able to ask some questions and have some interactivity with the panel panelists and the answer to that is yes absolutely. You should be able to ask any questions in the top right of your screen there's a little question mark box and you can type questions there. I will put those to the panelists in the second half of the webinar today. And then the second thing that often gets asked is, will we get the slides? Will we be able to catch up on a recording? Can I share this with my colleagues who weren't able to join?

Tilly Gilbert1:35
And the answer to that too is yes, just after the webinar and probably tomorrow you will get an email from us with a link to the slides and also to a recording of the webinar. So please do bear that in mind as we get on with the content of today's session. As I mentioned, the running order for today will be, first of all, a presentation from my colleague Miriam at STL Partners to set up the conversation. Then there will be a short presentation from Luis from Nearby Computing, and then we will move on to the panel discussion where we will be joined by two individuals from Orange Spain and also Hyon from Dell Technologies. But without further ado, first of all, I will be handing over to Miriam to kick us off with the initial overview presentation.

Miriam Sabapathy2:32
Thanks so much, Tilly. So yeah, just to introduce myself, I'm Miriam. I'm one of the senior consultants at STL Partners. And over the next 10 minutes or so, I'll cover how we see the AI Edge and private 5G opportunity for telcos and finish off with a couple of key recommendations for telcos. So, just to start with, just want to set the scene with a bit of what we see as a trend across the industry and over the last few years we have seen that AI has been a catalyst for Edge and private 5G deployments and specifically when I say AI, I mean a specific type of AI which is physical AI. And most people think of AI as software, things like a chat bot or an analytics tool hosted in the cloud or delivered over the public internet, but physical AI is different.

Miriam Sabapathy3:20
It's fed by data from the physical world, from sensors and cameras, and it interprets that data and acts on it in real time in the real world on a site. So think about things like robots, automated guided vehicles, camera-based inspection, autonomous trucks, et cetera. And that difference really matters here because physical AI has its own requirements. It has to react in fractions of a second. It generates a lot of data. That data is often sensitive. So the deployment model is more complex. And that's where Edge and Private 5G come in. And two quick definitions to make sure we're aligned about what we mean when we say Edge and Private 5G. By Edge, I mean, and for the purposes of this presentation, I mean compute deployed on an enterprise's own site for that enterprise's own workload.

Miriam Sabapathy4:10
So that can range from a small piece of hardware up to a full rack of servers. And by private 5G, I mean a 5G network that's dedicated to a site with its own radio equipment at sometimes sharing a core network. So we see that some key physical AI use cases are driving the need for both Edge and private 5G. On the Edge side, like I mentioned, there's a lot of data, sending that all back to the cloud is slow and costly. The AI needs to act quickly, so latency matters, and also sensitive data often needs to stay on site. For private 5G, many of these assets move in some of these use cases, like automated guided vehicles, for example, so they need continuous coverage with no gaps.

Miriam Sabapathy4:52
They need a higher reliability than a shared network offers and often higher security, which a dedicated network provides. And we see this being reflected in what we see from the growing enterprise announcements and deployments of private 5G. And I've listed here a few from our private networks deployment tracker in research. And we see that many large enterprises across verticals, such as manufacturing, to ports and transport, are deploying private 5G to support automation and AI-enabled use cases. And the details of some of these use cases I'll touch on a little bit later on in this presentation. And up until now, the industry and we ourselves have spoken about AI, and specifically physical AI, driving Edge and driving private 5G in slightly separate conversations.

Miriam Sabapathy5:40
But both are happening at the same time. So we do see demand for something a bit more converged. So seeing Edge and private 5G have two halves of the same proposition. And there are a couple of reasons for this. Firstly, shared but local compute. So you can run physical AI on the device itself, on the AGV or the camera. But most of the most valuable use cases work better when you combine data from many different sources. You know, a camera on its own can only really see its own feed, but if you offload that to a local server, you can combine that feed with your two other cameras along the production line, as well as things like temperature and vibration sensors and spot things that no single device can.

Miriam Sabapathy6:22
But of course, a camera doesn't have the compute or the storage to do that, so that's where we see a lot of benefit to AI needing both because as well once the analysis leaves the device you need a good connection to get to that shared compute as well. Second latency, so we see that Edge and Private 5G having kind of mutually reinforcing benefits for performance because the response time does depend on both the compute and the network. So a fast local server behind a slow connection doesn't help, so together they make latency lower and more predictable. And thirdly this theme around So for data to stay on site, both the compute and the network have to be local.

Miriam Sabapathy7:05
And sovereignty is the word that we hear most, and it matters. But often it goes a little bit further into things around corporate policy. So control and compliance for data that is sensitive IP needs to stay local. So either way, what needs to be local is both the compute and the network. And fourth, finally, and arguably most important, is the enterprise view. They don't want to buy two technologies. what they want is a working use case instead. So if Edge and Private 5G are both enablers of the same outcome of a use case, the customer wants one proposition, not necessarily two separate products. And we think where this matters most, where enterprises will want more of a joint proposition because they have this joint need for Private 5G and Edge is in certain verticals like manufacturing, logistics, and extractives.

Miriam Sabapathy7:55
But to make this a little bit more concrete, I want to walk through what this could look like in practice, thinking about firstly in this case, yes, smart manufacturing factory floor, and going through how the different physical AI use cases might work across the site and how they might work in tandem and where Edge and Private 5G come in. As you can see here on the slide, a bit of a diagram for what a smart factory floor could look like. On one end, you could have use case of automated guided vehicles. and to clarify these are driverless vehicles that move goods around the site and they are moving so they need to have a continuous connection across the site which could span indoor and outdoors so private 5g can offer that and if navigation is coordinated by compute nearby each vehicle then needs less intelligence on board which makes it also cheaper and then around the assembly line of the floor you could have computer vision for safety and security so that would use cameras around the site that use AI to check that workers are wearing the right safety equipment or to spot dangerous situations, or also for things like quality inspection as well.

Miriam Sabapathy9:04
And that requires a lot of high definition video, which the so the network does need to cover the whole site without any gaps. But processing the video on site also helps to cut the load on the network and keep the footage local as well if it is sensitive. Thirdly, a use case that we work across site could be robotic control of industrial machinery. And this is software controlling the machines and robotics in real time and correcting problems as they happen. The edge benefit is that machines can share servers in the plant rather than each device carrying its own compute, but also it requires very low latency, which is where both the private 5G and the edge also supports.

Miriam Sabapathy9:48
And fourthly, you could also have predictive maintenance working across the site. And this uses sensors on the equipment to measure things like vibrations, temperature, and pressure to spot early winds of wear and tear. So you can maintain the machine before it breaks rather than after, so more proactive maintenance. And it generates and processes a lot of data as a use case. So the private 5G network can collect it more reliably, even from the assets that are moving across the site. And Edge can help filter it on site instead of sending everything back to the cloud. As you can see here, these are just four specific use case once to call out, but you can see that each use case needs Edge and Private 5G to different degrees, but both of those technologies help in each case in improving the performance and can show as well about how physical AI, when it is working in the real life in real time, how they might be working in tandem for an enterprise as an overall link of their operations.

Miriam Sabapathy10:45
But how big could this opportunity be for Telcos overall when we talk about more of a joint proposition? And as many of you know, we have a private networks forecast and an edge computing forecast. And what we've done for this webinar is provided a bit of a combined view of what we think the converged opportunity might look like. And we've focused this on just three verticals where we think there's a real opportunity manufacturing, logistics, and extractives. And just to clarify what this number really shows is that the way that we model our forecasts, it's that it's the revenue associated with edge enabled and private network enabled use cases.

Miriam Sabapathy11:26
So not just the revenue from edge and private networking equipment. So as you can see, it includes revenues from things like devices and applications. And this combined view from just these three verticals covers the four use cases I just walked through, along with others that matter most in these industries. So things like computer vision for asset monitoring and quality inspection, mixed reality for worker safety, connected vehicle collaboration, drone inspection, et cetera. And as you can see, by 2030, we think revenue in just those three verticals could reach around $54 billion. And just looking at how those revenues are split across those different layers in 2030, around 70% of the opportunity is in device and applications.

Miriam Sabapathy12:09
So the layers that a telco is more likely to address around private network connectivity and edge platform infrastructure and as well, potentially integration and support. So a telco that only plays within the infrastructure layers, that being private networking and edge platform addresses around seven billion dollars. But if it takes more of an active proactive role in the integration layer that could rise you towards 15 billion. So where it lands depends on how far up the stack a telco might go and on winning that, that would require you really competing against some of the systems integrators. But what we do see despite 70% of the revenues being in devices and applications, the edge and connectivity layers are also the fastest growing even though it's working from quite a small base in the earlier years.

Miriam Sabapathy12:59
So what we wanted to show here really is that even just looking at three verticals, there is a real case for an edge and private 5G converged proposition and a material opportunity to address.

Miriam Sabapathy13:13
So just to finish off this presentation with a couple of key recommendations for telcos. Our first one is really about focusing your efforts on verticals. If you do want to offer a converged edge private network and AI proposition, the verticals that need it most do require industry-specific know-how. use cases are vertical specific and enterprises need to trust you as a partner to understand how they work and what their challenges and priorities are. So rather than a broader horizontal offer, we think that growth will come from specific propositions in specific verticals. So make a call about where you're going to focus. As we mentioned, where we think the biggest opportunity lies is manufacturing, of course, but also logistics and extractives as well as shown on previous slide.

Miriam Sabapathy13:56
The second is about positioning and we our recommendation is think about the orchestration opportunity and think about using the network and the edge as one platform. You tell us your positioning, you're very well placed to sit across connectivity and compute your more unique positioning than others in the ecosystem, you can't be replaced there and taking that role of orchestration helps improve what the enterprise gets around application performance, your security, latency and with it has that mutually reinforcing benefit which matters a lot for mission critical AI and I know this in particular is what Luis will be covering in his presentation.

Miriam Sabapathy14:33
And finally, it's about speed and urgency. I think the real recommendation here is move sooner rather than later. We are seeing that physical AI is really starting to appear in announcements from enterprises, from ecosystem players. We're seeing a lot from our private networks deployment tracker about a lot of the deployments being linked to AI and automation led use cases. But also starting up a new telco proposition takes time, you know, first trying to build the business case internally, and then longer still to actually get enterprises buying and reaching revenue. And in the meantime, specialists and SIs, we do see our gaining share, especially within the private network space.

Miriam Sabapathy15:10
We're seeing that a lot of those specialist systems integrators are much stronger in the last couple of years. And some of them as well are starting to package specifically physical AI. So for example, NTT Data and Ericsson have launched a partnership this year thinking about a private 5G edge and physical AI combined proposition. So I'll close there with the end of our presentation. So now I'll hand over to Luis who's going to talk a little bit more about the orchestration opportunity and the experience in practice. So over to you Luis.

Tilly Gilbert15:55
Luis, are you able to hop on over and drive the presentation? There we go. Sorry, you're just on mute then, but I'll hand over to you.

Luis Zas16:06
All right. Thank you very much. Thank you, Miriam. Good morning to everyone and thank you to STL for having us on this panel. Part of what I will share today comes directly from our work with or into Spain on a new product named Connecta 5G. Before we enter in the details on the product, let me walk you through the overall concept of what the product looks like and what we are doing here. So the title of the presentation is Converging Connectivity, Edge, and AI. And that's exactly the journey that I want to walk you through what on the one hand, we're gonna see how the telco edge cloud is where the network, the cloud and the AI finally meet. And on the other hand, what an operator needs to put in place.

Luis Zas17:09
So that convergence turns into revenue rather than into complexity. So if we move on to the next slide. Thank you. So, after more than 25 years in the industry, both on the operator side deploying services and networks, and on the other side on the vendor, I can tell you, the questions has not changed at all. So, three of them keep coming back. And we are going to present a couple of slides through those questions. So the questions are, how much money I can make? Or in other words, how can I open new revenue streams beyond connectivity? The second question is, how much money I can save? So how can I keep my operational expenses flat while the network, the cloud, and the number of sites keep growing.

Luis Zas18:19
Last but not least is how can I differentiate myself from my competitors? So not only the operators, also from the hyperscalers that are trying to move towards the edge. What has changed though, is that the answer to these three questions goes through the same decision. How do you orchestrate? Orchestration is not longer a technology choice that you make after the architecture is settled. So it's the foundation of all what you do is sustainable and define the edge strategy. The operator that orchestrates well, launches services faster, operates more efficiently and can expose the network as a product. The one that doesn't ends up with a collection of silos, one vendor, one per domain, and someone else will capture the value of that edge.

Luis Zas19:28
So new revenue streams are emerging across edge, cloud, private networks, and AI. And operational efficiency is becoming a strategic differentiator rather than a cost line. So on the edge and the proximity, sovereignty and control creates a position that the telco only holds. So the message on this slide is, the message to this slide is telcos must move faster or someone else will capture the value. So let me tell you through the three questions. So the revenue, so as for how to increase revenue, both by providing connectivity services or by unleashing the cloud native capabilities within the telco, monetization starts when the time to market moves from months, even years down to the weeks.

Luis Zas20:33
So we see four pillars that underpin the revenue creation. The first pillar is the traditional connectivity, fixed and mobile. Telcos remain the master of the connectivity. So because they have the infrastructure, they have the experience, they have the know-how, so they are still keeping this revenue stream as the foundation of the revenue. However, now we need to keep adding more new services. Traditional cloud services, the first one. IaaS as a service pass must be part of the portfolio as well. We've been talking about containers as a service for years, but the truth today is that we are still seeing requirements to manage both virtual machines and containers in a multi-cloud environment, both private and public.

Luis Zas21:28
So the third pillar, new services with a short time to market. So new applications that complete the telco offering, and meet the B2B customer are asking for. So here, the only thing that we can say is, be ready for the unknown. Going forward, AI will either be infused into the operational workflows that we are seeing already, or offer to the end customer like other applications. and the platform that you build has to be ready for both. The fourth one, customer empowerment. Give more control to the customer. Self-managing through end-customer portals, where the customer can activate and deactivate services without opening a ticket. They can purchase through catalogs and marketplaces.

Luis Zas22:26
So the same portal that can activate a cloud workload, let enterprise manage their own private network. If we move on to the next slide, please. On the operational expenses, this is a really interesting one. The moment we say, unleash all the capabilities of native and change the paradigm of how multi-cloud networks are operated, customers start hearing one thing, and it's more complexity in my operational processes. This is where a good orchestration strategy plays the key role. Two requests that we are hearing from the customers as to how to the OPEX flag or to have OPEX efficiency. The first requirement, zero touch provisioning, fully automated provisioning of the services and integration of the network function from the edge to the site and including the public cloud.

Luis Zas23:47
So with the operating declared in the intent and the platform convergent to it. So the second request is a single pane of glass. So we are talking about events, alarms, KPIs collected and correlated in one place, regardless of what is running underneath, private or public or whoever the vendor is. So these two requirements apply to Cloud, but also to the private networks exactly in the same way. So if we move on to the next slide, the third question that we mentioned, how we can differentiate ourselves from the competitors. Here, we need to talk about the edge. We define the edge as anywhere close to the end customer, including the customer premises.

Luis Zas24:45
Here, I have to say that nobody knows how to manage the edge, better than the telco. So you know the power constraints, the space constraints, the field operations. So nobody else has the capability of the infrastructure that the telco can deploy. Since 2018, we've been talking about the edge as an enabler of low latency services. Today, the real drivers for deploying services are different. So basically are, and Miriam mentioned something earlier. So sovereignty, understood as sovereignty within the country, regional level, and sometimes even on-prem. So the second driver, data protection and security. Data must not leave the plant, the port, the hospital, with security enforced locally and by design.

Luis Zas25:53
And the third driver is tracking efficiency. Some applications process so much data that is simply more efficient to host them in the way the data is produced. So less by whole, less bandwidth cost, better economics overall. So I would add an afford element here and is the network control guaranteed bandwidth and SLA become as part of the service as well. And this is what it turns the edge site into a differentiated offer as well. So moving on to the next slide.

Luis Zas26:41
So now we have to talk about Connecta 5G. This product, Orange Spain, launched in April 2026. I have to say that this is the first all-in-one platform in the country where enterprise and public administration manage their private 5G network and they make infrastructure underneath. from a single pane of glass across multi-vendor private networks and multi-cloud environment that we include private, public, and on-prem. So you can see in the slide that how Connecta 5G is built. We see three legs that reinforce each other. The first one, the first leg, private network connectivity. the 5G network itself on the public core as an entry product and a standalone solution as the premium one, where the customer has the possibility to manage the slices, the DNNs, the SIMs, the CPEs from the portal.

Luis Zas28:01
This is the first leg. The second one, of course, edge cloud, make public clouds, private clouds, and on-prem clouds, where the applications can run next to the data. The third leg, operational efficiency. So instead of having different many portals, different tools to manage and configure and monitor the network, we have all in one zero touch operation and fast adoption. So if you take away one of the legs, the other two will lose. And that's why they are on the same platform. So connect FID is the model. One portal on the public and standalone and control for the enterprise. If we move to the next slide, please.

Luis Zas29:00
So, and now what? because you have the edge cloud, the services at the edge and evolving towards AI. I have to say that the next step of the telco is not more IaaS or PaaS services anymore. It's AI. The telco already has the sites, the power, the connectivity, adding accelerated compute to the footprint turns the edge into a mini factory close to the customer. GPU as a service is the next revenue line that we've seen and we expect that rests on four ideas. The first one, as we mentioned, GPU as a service. Accelerated compute added to the sites, the telco net, they cancel it. They have the infrastructure already. The second one, the second idea, inference at the edge.

Luis Zas30:05
Models run where the data is produced. So less backhaul, lower latency, and better economics. The third one, sovereign AI for enterprise. Data models and data and models in staying country, in region, and on-prem. for customers that they cannot send the data to the hyperscalers. The last one, and this is a conclusion of what we are talking about, is the same site, the same platform. The site that is hosting the private network can also host the CPUs that process all the data. Deploy Zero-Touch and operate it exactly the same way. If we move to the next slide, bottom line. So the most important part is define a long-term orchestration strategy that will help you get the most of what the orchestration can bring.

Luis Zas31:14
That's the first thing. The second one, improve the time to market of your services, moving from months to weeks. Third, streamline the operation processes. Zero touch provisioning and single pane of less whatever is running underneath. The fourth one, and it's more focused on the private networks, two different flavors of it. The public core and also the standalone solution operating on the same platform. The fifth one I would say is building open architectures with a strong ecosystem. So be ready for the unknown. And last but not least, is doing nothing is an option because the opportunity to grow in a B2B revenue is already there. And AI will come next. Moving on to the next slide.

Luis Zas32:16
And thank you very much and very much looking forward to to the panel. Back to you, Tilly.

Tilly Gilbert32:16
Thank you, Luis. Thank you so much. And can I welcome the other panelists to join as well?

Tilly Gilbert32:32
Turn on your cameras and be ready to be part of the discussion. Thank you. Hi, Hyon. Perhaps we'll start with you and give you a second, Hyon, to introduce yourself and your role at Dell. And then We will also do the same for Rafael and Pedro before we jump into some of the questions I have. Also, just to remind the audience that feel free to submit your own questions in the top right-hand side if there's anything that you would like to ask the panelists today. But, Hyon, maybe we'll start with you with a quick introduction.

Hyon Dijksterhuis33:08
Yes, Hyon Dijksterhuis is based in the Netherlands, director in EMEA at Dell that's focused on telco and working with our telco partners grow their B2B business. Thank you. Rafael and Pedro, will you also introduce yourselves from the Orange Spain side?

Rafael González González33:20
I'm Rafael. I'm the head of marketing large accounts products in Orange Spain.

Pedro Herbello33:27
Good morning everyone. I'm Pedro Herbello. I'm a senior product manager for 5G product line here in Orange Spain.

Tilly Gilbert33:44
And maybe I can start with you both on the Orange Spain story because Luis mentioned a little bit about the work that's been done on Connecta 5G in his presentation but I would be interested to hear from you both around why Orange Spain built Connecta 5G and what the enterprise customer challenges were that you are looking to solve with that product and that proposition launch. So perhaps you can tell us a little bit about the reason behind the launch of Connecta 5G and what it is you're hoping to solve for?

Pedro Herbello34:25
Yeah, we started as Orange Spain, we started deploying a private 5G network since 2022. So more than four years ago. At the initially every network was a bespoke project. So, we designed it and delivered it case by case. Okay, that worked a while, where the number of opportunities was small, but as demand grew, it was no longer sustainable. So we decided to produce our private network offer. So we launched it in 2025, our 5G MPN products. So after that decision, we saw that the number of customers was increased a lot. I mean, so the number of customers grew too much.

Tilly Gilbert35:36
Too much, no, it's never too much, so.

Pedro Herbello35:38
But at the same time, the number of customer increasing, some of our customers started to ask more than connectivity. So of course they wanted a guarantee of connectivity, but they also needed MEC and private cloud to run their applications close to their data, as Luis mentioned it before. So we had a growing base of private networks plus edge infrastructure, each with its own tools and processes. So managing that 5G MPN ecosystem was getting very complex, mainly for our customers. So that's the point when we decided to build our Connecta 5G platform as a single self-management platform that serves two audiences. of course, for our customers, so they were able to manage their own network and edge services, such as SIMs, Slices, Devices, and Cloud, without depending on us.

Pedro Herbello37:00
But also, we made it for our internal operations and maintenance people. So both our people use the same tool to operate every deployment. The result, it made a simpler operation, reduced the customer tickets and lower our operation cost.

Tilly Gilbert37:28
That's very interesting. What you spoke about and the implication behind that was that you were seeing enough demand for the private networks and then the additional edge environment for it to be worth your while to standardize, to make it internally more efficient to deliver those networks. And so I think, of course, we would be interested to hear what you can say about where that customer demand is coming from either from an industry vertical perspective or perhaps from the perspective of the types of use cases or applications that they were typically coming to you guys with and saying we need a more dedicated, more specialized network and compute proposition to help us to be able to deploy these types of applications.

Tilly Gilbert38:29
So, the implication is that there was really good strong demand out there for you guys to go on this Connecta 5G journey and I'm intrigued to learn a little bit more about what that demand looks like and how it's changing at the moment.

Pedro Herbello38:49
Do you mean what we foresee for the… What types of customers have been driving the most conversations at the moment and do think that will change going forward or the same types of customers? Yeah, the main driver is clear. Use cases where the wireless connectivity is critical for the customer's operation. So this is where we started. So for us, a milestone was the Port of Barcelona MPN. So, with that project, with the success, the success in that project gave us a lot of experience and knowledge, which we used to push commercially into that vertical, I mean, logistics and port sectors with large size outdoor operation and a lot a lot of moving assets.

Pedro Herbello39:53
So at the beginning, we pushed our sales force to that two verticals, ports and logistics, which I would say is quite a success. So we have many other customers in that strong demand in manufacturing and industry, of course, where the reliable wireless connectivity is becoming a requirement for automation and connected operation, as Miriam mentioned before. So regarding the future of Connecta 5G, I see AI is for sure in the equation. The next logical step for Connecta 5G, of course, we plan to include AI in the platform. We are starting right now, we are evolving the platform and we are starting to work on that because we think that AI is going to give us a differential value for the customer.

Tilly Gilbert41:17
And here on then, Luis, you guys work with lots of other telcos as well as Orange Spain. And so I'm interested to hear from you guys whether you're hearing very similar things from the other telcos that you're working with in terms of kind of the drivers and the use cases and the customer demand, or whether there's anything that either of you two would add from your own experience in terms of what you're seeing in the market?

Luis Zas41:23
So thanks for the question first of all. So yes, definitely we are seeing some common denominator across the demand. So we are seeing that one of the main drivers as Pedro said is, okay, I want to optimize, I want to optimize my infrastructure. The way and the reality is that some customer dictates the type of bird that they want to use and that add more complexity and increase the OPEX and the operational expenses for the operator. So as far as the European demand is concerned, we are seeing two different threats. The first one is really focused on, as we said, mobile private network, and also how to leverage the current infrastructure to become a mini AI factory.

Luis Zas42:36
turning the cloud edge into a factories as well. There are different threats that are happening now in Europe. The common denominator for those are first, increased revenue opportunities through the existing connectivity, the traditional method. But also on the other hand, trying to see how I can leverage that infrastructure, and turned that infrastructure into a new potential revenue stream as well by using cloud closer to the customer or AI in the future as well with GPU as a service.

Tilly Gilbert43:19
But to the point, Luis, and this is a conversation we are having with several of the telcos, I do think, Pedro, the point that you made when we're looking at customers, they are really looking for guidance in a lot of cases around what are the use cases, What are the outcomes that are going to help me? And in the case of the ports, the challenges as in having real-time insight, when what comes in, how many employees do I need to have on site? Where is the safety security risk? How do I alert the right medical staff in order when a medical incident happens? All of those conversations with customers force the telco also to change their story and their thinking if they want to go beyond connectivity.

Tilly Gilbert44:02
And if we look at the solutions that are driving those AI supported outcomes, we do need to think about the fact that not one single player in the market can do all of that by themselves. So it is about building an ecosystem of a solution that depends on players coming together and building that together and the telco being central in having that relationship with the customer and being that service partner to the customer. And that's where it does require a lot of time invested upfront in thinking through that whole end-to-end solution process. And that's why we're all here at the table together, because all of us do need to come together in order to think through that solution that customers are willing to pay for.

Tilly Gilbert44:51
And Hyon, I think both Pedro and Luis mentioned AI as being a key future driver for Connecta 5G and for some of the work that they're doing. It obviously has an implication, there's been lots of discussion about the implications of running more AI workloads at the edge in terms of what that means for the hardware, for the servers that you guys are providing as part of the play for Dell. I'm interested in your reflections on that based on kind of the presentation that Miriam gave where she was speaking through some of the use cases. What are you guys observing and what are you recommending to customers in terms of how specifically AI at the edge might be shaping the hardware needs of enterprise customers?

Hyon Dijksterhuis45:52
And I think indeed that unfortunately there's no one single answer and one single solution that addresses everything. I do need to stress that we need to accept the fact that we really, if we take the customer interest to mind, that we need to realize, we need to look at what are we trying to accomplish. What is the best place where indeed that solution or that infrastructure is placed and that's where that answer is not always the same. Sometimes it is on-prem when indeed data cannot leave the premises. Sometimes it is a centralized location when I'm thinking, for example, about hospitals that could share within a group of hospitals a shared environment closer to the hospital or in other instances you can centralize it in the telco site.

Hyon Dijksterhuis46:42
So we do need accept the fact that when we also look at the platform implications, you need to accept the fact that you have that distributed approach that you need to support. But I think indeed, key is that you need to accept the fact that sometimes it's on a location where there's no data center, where there's no cooling. So product wise, that needs to be able to be supported. And we're able to do that. But also the fact that standardized approaches help, but accepting the that the locations might be varying and different. And to Luis' point, you need a common denominator in order to orchestrate, manage, and monitor all of those different locations and platforms in a single approach.

Hyon Dijksterhuis47:26
And I think that that is why we are working with Orange and Luis also on the fact that that's where we are accepting the reality of we need a hardware approach, but it needs to support customers' requirements on. Where's the location? Is it centralized? Decentralized.

Luis Zas47:44
I agree. If I can add here, what we are seeing, and this is across the board, I can see two main drivers as far as to AI deployment or demands and what are the drivers. The first one is governance. Whatever is going to happen, governance is going to be one of the key drivers as far as the AI deployment is concerned. The second one is cost to definitely there are companies that are investing now and GPUs and however We are going to see us as high on just set just mentioned We are going to see a hybrid environment as well What you're going to see and you're going to deploy? Workloads across different different environments on-prem and telco or even at a hyperscale trying to leverage all the infrastructure and

Tilly Gilbert48:23
And this is really a question to all of you, I think, or something to discuss a little bit further, because I think you all in some way mentioned this tension in these product offerings between trying to standardize to make it, you know, to make the telcos be able to move faster to, you know, work in as a blueprint as part of an ecosystem together.

Tilly Gilbert48:58
But then, Hion, you were just pointing out, you also have to be prepared to be bespoke to meet the customer needs where they are, to understand that each deployment, particularly for something like a private network, is quite specialized to the enterprise customer and their specific needs. So these two things, I think, are, they sit in tension with each other. So logically, yes. It's difficult to get the balance right. Can you talk a little bit more about that?

Hyon Dijksterhuis49:28
But I do think, indeed, even if the fact that you are accepting the fact that there's mainly three options, it's on-prem at the enterprise site, it is nearby the enterprise centralized or at the telco site, those are roughly three configurations that you can blueprint in practice that kind of approach. It's just that when it comes to combining it with an ISV application or solution, or looking at the scale, because it could be like three, four hospitals, could be also 50 manufacturing sites across the globe. That is where you need to cater and adjust the sizing and the platform approach to cater for that. But architect-wise, you can already accept the fact that you know roughly these three flavors that you can blueprint.

Hyon Dijksterhuis50:20
in order to get you started in a way and then adjust when you get to the details of what is it that the customer really wants to do.

Tilly Gilbert50:31
And Pedro and Rafael, from your guys's perspective, get this balance right between the more bespoke integration work, you know, consultative selling with your customers and something that is scalable and that you're able to, you know, deliver in an efficient manner. How are you thinking about that and, you know, more generally, what do you think that telcos like yourselves and your peers in the market, what do they need to be doing to be an effective channel for these sort of edge and private networking type solutions?

Rafael González González51:10
Well, just to answer your question, the first thing is that we have to manage both things. We have to be able to do bespoke projects, projects specifically designed for the customers and at the same time, the productification, the standardization of the solutions, which has to move forward in a very scale and a very productive way. So the first thing probably is, and that is what we learned in the past years, is the productification and standardization. We are not able to do everything as a telco. We need an ecosystem. We have to rely on the partners, but we need the a lot of products with a lot of things, well thought, well designed, just combining in public core and private core, just combining everything in a very simple way to deploy the solution.

Rafael González González52:21
The second step probably is to enable the sales force to sell not technology, but solutions. And this is the most difficult thing because we are telco operators, we are very used to sell connectivity, mobility, connectivity and everything, but the way we deploy mobile private networks at the same time age and at the same time AI is not a simple way to do. So try to sell, to assess the customers, not technology, but explaining and understand with a lot of accuracy the needs of our customers. The first, let's say pillar is to lower the barrier to entry. So we have to learn, start small, prove, fine tune and then explode and then grow with a very large scale.

Rafael González González53:26
So we have to go step by step and at the end, of course, work a straight ecosystem. This is the, I think, the most important value that a telco can provide. Connectivity opens the door to the data. The edge provides the place to process the data and the AI, which it seems today, the final step unlocks the value of everything. So this is a long trip. The operators, for sure, and we as Orange Spain with Connecta 5G, we have been pioneers in Spain. We have been the first operator selling these standard solutions. We need to build up a new ecosystem coming from connectivity and going forward to the AI. And in the middle, of course, combining the cloud, the security and everything that we manage in our networks, but for sure we need partners, we need to have the best partners with us just to employ the best solution to our customers.

Tilly Gilbert54:38
And I think that's really interesting, I'm intrigued in the final few minutes to maybe look forward to what's coming down the line, what you expect over the coming few years, particularly because I'm sure there will be people on the webinar who are thinking private networking as a concept is not necessarily new, you know, there have been enterprises that have deployed and managed their own, you know, networks for a long time now. Some of the customers that you guys are talking about, like the say ports are very real, but they have probably been the poster for private network deployments for the last few years as kind of, you know, they are definitely an environment that benefits from this type of dedicated network.

Tilly Gilbert55:28
But there are also only so many ports in the world and you guys are presumably thinking about a growth story I think that will go beyond, you know, that vertical and that type of customer. So I think Rafael, what you were saying is that the private networks and the edge environments offer an environment where AI use cases, particularly I think the physical AI use cases that Miriam were talking about, can deliver real new value to the enterprise customers, and that's part of the expected growth story for the next few years for Connecta 5G. But I'm interested to hear anything else that any of other panellists have to say on where we expect this industry to head next and what we expect to see over the coming few years from a customer needs and demand perspective particularly.

Hyon Dijksterhuis56:29
I think that Miriam already mentioned in the beginning it is about if you want to capture the market you do need to be perceived as the player or the party at the table that have have expertise and have value to add and I think indeed that verticalized use case approach is what customers will be convinced by in taking that AI step because there's so much out there in the market. A lot of companies are piloting but it is the hesitation from pilot to production that is stopping the real acceleration and I think that that's where building enough expertise, getting that vertical relevance and leveraging, particularly for the telcos, leveraging the footprint they already have with their connectivity background is key to leverage and understand what those customers are looking for.

Hyon Dijksterhuis57:24
But they are looking for a single player that leads that conversation and that brings an ecosystem of partners to bring that AI-relevant solution to life.

Luis Zas57:37
So from my experience so far, speaking with different customers where I'm seeing is, the first one is trying to make sure that the underlying infrastructure is ready to put in any particular sector. Then the sector demands particular applications, but as long as you have an underlying infrastructure and architecture that can support different applications and there is any impact on the time to market, I think you are really well-equipped to cope with the demands from different sectors. It's true that depending on the sector, I'm talking about with smart cities, how they are ramping up now, so we can see that there are a specific needs as to the applications they're going to use.

Luis Zas58:31
But the most important thing, and I see exactly Connecta 5G is a really good example is the architecture is ready for that. So then they need to enter in exactly what applications the customer are demanding. So I think is the most important part here.

Rafael González González58:48
Yeah, I fully agree. At the end, we realized that customers did not just need a private network. They needed a simple way to manage their entire private connectivity and edge environment with the same level of self-service, visibility and control they expect from public cloud platforms. So they know the hyperscalers solutions and they expect from this new private networks plus edge, plus AI, the same level of reliability. So at the end, they were looking for simplicity, visibility, control and agility. and this is the game of Connecta 5G and the way we are going to improve these solutions to the market. This is a long trip and we are just at the starting point, so it's just as exciting for everyone who is involved on that.

Tilly Gilbert59:47
Well, I can see that we are at time for today's session, but I think some of those final parting messages about, you know, Hyon, from the kind of the need to verticalise and to really understand in a few specific areas what those enterprise customers need, I think is a great message. I think this message around the need to create something that's scalable and repeatable behind the scenes while also bespoke to what the customer wants is also key. And then Rafael, your point at the end there about it's got to be easy for the customer to consume and use and control, I think is also a great one to leave people with. So thank you all for joining the webinar today.

Tilly Gilbert60:33
Thank you to my panelists for their contributions. And yes, we will share around the link and the recording and the slides to you all shortly. Have a lovely rest of your day. Thanks everybody. Thank you. Thank you.