Amdocs Limited (DOX) Earnings Call Transcript & Summary

Hi everyone, and welcome to this 1-hour investor webinar, the journey is to Streamline and Automate Network Ecosystems in the 5G and Cloud Era. I'm Matt Smith, Head of IR for Amdocs. Before we get going, a reminder that today's discussion will include forward-looking statements, which are subject to the risks and uncertainties outlined in our SEC filings, we might also reference financial measures that are non-GAAP in nature. The reconciliation of which to the most comparable GAAP provisions can be located on the Amdocs IR website and also in our 6-K filings with the SEC. The purpose of today is to talk about network automation, which is one of the four main pillars of Amdocs strategic growth framework. To help us with that, we have three expert speakers for you today from Amdocs Networks, Niall Norton, Division President; and Ana Redondo, Product Strategy lead. We're also privileged to welcome Gint Atkinson, VP Network Strategy and Architecture for SES an Amdocs Customer and one of the world's leading satellite communications providers. In terms of the agenda, Niall will discuss Amdocs' end-to-end network automation journey, and it will then drill down on how Amdocs is meeting the network automation need. Following which we'll sit down for a virtual fireside with Gint to discuss SES's network automation journey to date and its partnership with Amdocs. And to the extent we have time left, we'll take some Q&A. Questions can be submitted over the chat feature on your web browser. Now before we get going, let me provide some context for today's webinar. As many of you know, Amdocs provides mission-critical software and services to more than 350 of the world's largest CSPs including the likes of AT&T, T-Mobile, Val, Vodafone, SingTel and SES, of course, we enable CSPs to design and create new services to monetize those services and also to provide a great experience to their own consumer and enterprise customers. And we do it by bringing our unique model of software deployment capabilities and operations. The combination of that resulted in very long-term customer engagements for Amdocs and highly recurring revenue streams. Now over the last few years, we've accelerated our growth by expanding our serviceable addressable market which we think will actually approach around about $57 billion by 2025, that's up from roughly $36 billion in 2020. This expansion of some reflects our strategy to really bring our customers the market-leading innovation they need to support the next-generation technology investments they are making in across several key domains. These include digital modernization, which was and remains a core growth engine for Amdocs, both pre and post pandemic. But since 2021, we've also added three other growth engines, journey to the cloud, 5G monetization, where we're enabling CSPs to deliver the exciting new 5G services and use cases that we expect to be made in the coming years. And related to that is network automation, which is the growth engine we want to focus on today, given that it's a natural flow one, if you like, from the 5G webinar that we brought to you last December. And also that we believe Amdocs is a very unique opportunity and emerging role to play as networks become a lot more virtualized and software-defined in the future. So with that, I'm going to be delighted to introduce Niall and hand things over to him. Niall, please take it away.

Thank you very much, Matt, and welcome, everybody, to the webinar. My section, I'll deal a lot with the kind of business strategy aspect of what Amdocs are doing in the network automation space. I'll kick off, if I may, with the -- just a little bit of context around the industry. Starting with Amdocs has been actually in the network domain for about 15 years at this point in time. Now network automation has undergone a gradual evolution over that period. We have been very successful in the past in pushing the IT domain back down into the network in different places. But we have a very good pedigree. We've done a series of acquisitions in the space, very surgically targeted towards places where the value can be added. But the -- so our experience in this space isn't new. What we're seeing now is some changes in the industry, which are causing considerable changes in the business model for CSPs, both on the cost side of the equation and the monetization side of the equation. So without delaying too much on anything, if I jump to the next slide, I just want to provide a little context to the era of ubiquitous connectivity is phenomenal, and it's great to provide a great deal of innovation for what CSPs can provide. It comes with a degree of complexity of new technologies because a lot of the opportunities that we're seeing now, not beyond 5G, the enabling technologies coming from AI and ML for multi-cloud environments, from network slicing, new services like cloud gaming, metaverse and so on. These things are enabling technologies, which for the first time have really come together to allow true ubiquitous connectivity come to pass. And I think you'll have seen that. Today's focus is very much around the journey of automation and what that will mean to the operators' businesses. And Matt, if you don't mind going to the next slide, the first and most present opportunity for the CSP community is having spent 30-plus years evolving a consumer network and a consumer network includes kind of business uses for communications, that -- those services are broadening out into a much richer vein of different useful services that the operators can touch on to. But at the same time, because of advances in cloud and edge and other things, we're seeing also a new market emerge for CSPs or in the enterprise and industrial network where robotics and smart cities and things like that are also coming about. That brings with it some challenges to the operator, not only technology. But if you jump to the next slide, Matt, you'll see a range of new challenges also in the business model that the telco operators operate and how you engage to monetize different revenue opportunities requires an enabling technology at the back end, which is quite new to a lot of CSPs and the systems required and the business process is required to enable those, the flexibility required in the network, it's a transformation. It's not an incremental move -- and it's a multiyear transformation, which will go on for a decade plus as telco becomes more embedded in the fabric of everyday life and ultimately, everyday business. The second kind of, I guess, big challenge beyond the kind of business model is the speed at which new services are being adopted. And if you go to the next slide, Matt, I just there's some -- one of the things we track is the rate of innovation on the velocity of innovation. And you see in the slide, CSPs must operate in a complex ecosystem for sure. That rapidly changing market dynamics and exponential technology shifts, they're speeding up, and they are quickly evolving the human adoption behaviors, how do you train people, how do you have customers who can adopt your services? But that velocity of adoption of new services is rapidly speeding on. ChatGPT got to 1 million users in a day. This was incrementally faster than some of the more modern services have been launched. So not only have CSPs the opportunity to participate in a much expanding market capability with a much greater range of services in each of those markets. But the speed at which those services are being adopted is also moving really, really fast. So fast launch time lines, the explosion of devices and service launches. And most importantly, the cost effectiveness of being able to deploy the services and manage the network that's actually hosting those services. These are big challenges for the CSPs. And if we move on to the next slide, please, Matt. You'll see what we try to capture here is the degree of the multiyear journey aspect of this because it's not done in isolation of existing networks, it's not done in isolation of any individual technology trumping anything else. The scale divisions of the devices becoming live native and developing the edge of the network at the cloud being able to orchestrate connectivity across wireless and wireline and satellite technology platforms is a big, big thing. A lot of it focused on the business outcome. So as the operating paradigm moves from optimization of the rand, the embracing of the network edge, the adoption of ecosystems the extreme automation in what people want to use this technology for. These are big challenges that the operators have. And this is a multiyear journey that we've been embarked on with our customers. For Amdocs, it's a phenomenal opportunity because where we've been a player predominantly in the IT side of businesses. These changes are bringing about transformational requirements in the network side of the business, probably for the first time ever at anything like the scale that people are seeing. And Matt, if you go to the next slide, I appreciate I'm moving through these fairly quickly. I just want to touch on a couple of key points. But the automation is inevitable in growing industry. So what we're talking about today and the growth of automation and the rapid speed of automation in telco is already a play in multiple other industries. The pictures are kind of very much talking to the kind of almost exponential growth in the requirements for automation because for the last 100 years, it's been moving and moving faster all the time, and we're now entering into this 2023 and beyond. Moving into extremely complex networks, which are required in order to deliver the service capabilities. The thesis for what the operators are actually trying to deliver. If you jump to the next slide, the nice part of this is, it's not technology for technology's sake. A lot of the automation journeys that the operators are on is what we call intelligent automation. It's a big delivering business value. So it's not invest and they will come, automation becomes a must-have for the creation of new monetization opportunities, accelerating the time to market, exposing the network to the partner ecosystem, very, very challenging to try and do this with existing processes and existing people-heavy systems, it's almost impossible to do without having edge capability in the cloud. So on the improving the TCO and keeping all of this to be able to be done in a cost-effective manner, improving the business and operational efficiencies modernizing existing platforms, process and services, and enhancing the service infrastructure evolution. This is what intelligent automation can do. And that is -- and I will -- my colleague Ana will touch on this in a little more detail. But everything that we're seeing, everything that we're engaged in with our customers is significantly driven off business cases. It's not, as I say, a technology and they will come. Now on the next slide, we just talk a little bit to how far along is this. And our experience, which I think is mirrored by a lot of the analysts experiences, that the telcos have all -- understand the requirement to automate. They all understand the benefits in the business case. At this point in time, this is nascent. As a market opportunity for Amdocs, it is in the early stages of what it will be because this is a multiyear, very significant transformation in every CSPs business. There's some interesting kind of statistics coming out of these [ monthly ] reports and other analyst reports that we see, where machine learning and the adoption of automation is still in the early stages, and there are reasons for that. And if we go to the next slide, very pragmatically why is this -- why is it an opportunity for us, but why is it moving more slowly up until now? A lot of people will say, gosh, we've seen full stones in CSP industry before. But the reality behind the -- what I'm going to say, the slowness up until now of this being adopted, why isn't it here already because cloud isn't new and so on. It's very much that the operator environment is very complex. CSP need to support multiple existing platforms and adopt new ones, 5G and others. And that's a heck of a juggling act to actually keep up with. New technology and tools, networks are evolving in the [ 5G ] era from being physical boxes to being virtual software platforms. And those software platforms are operating, not in single unit centralized data centers, but those capabilities need to be at the edge as well as on the cloud with virtual services. So if you're doing a lot of data-intensive processing with billions of devices at the edge of the network, it makes no sense to try and transport that traffic and then the intelligence all the way across the network to simply move it back down again. So adopting new technologies and integrating those with existing systems is a very complex challenge. Most people tend to overlook a lot of the people aspect of this. Every operator has significant existing investment in training and people and processes. And the move to virtual assets rather than physical assets obviously involves changes in technology, but it involves changes for all of those people as well. It's a very significant change management program required around all of that. If this was simply a world of greenfield operators, things would move probably more quickly, more elegantly, but that's not the way the real world works. And so ultimately, what we've -- what our experience and what we believe is held out for the analysts readout so far is the speed of change is happening, but CSPs really require help to successfully execute these transformation programs. And that's -- if we move maybe to the next slide, and it will be my last slide before I handover to Ana. But just from a business perspective, our opportunity, and this is an opportunity that we believe gives us unique differentiation versus others in the industry. We don't see an opportunity where Amdocs is trying to compete with an Ericsson or Nokia, those people, they provide capabilities which we're respectful of. But what we see is this transformation, this move towards automation. What we bring to the table is best-in-class products and ecosystem of relationships. A massive experience between from the BSS all the way through -- the billing all the way through to the edge of the network. We have all of that experience. But as importantly, we are recognized as being an incredibly safe pair of hands to bring people on a journey. We understand the politics change. We understand the people impact which have changed the business processes required around that as well as being agnostic as regards software tools we bring. So regardless of which nets you're working with or which cloud providers you're working with, we're able to bring the expertise across multiple domains. We're able to bring an agnosticism to whatever network type you have. To become -- to allow the operators to really become agile and embrace intelligent automation as quickly as possible. I apologize if I move through this very quickly. It's the first of being an Irish guy is speaking too quickly, but I was hoping to make sure that I left plenty of time from my colleague, Ana, who I will hand over now to talk a little more on double taking and some of the technology side of things.

Thank you, Niall. Thank you very much. Thank you all for your time. I'd like to start by explaining what is network automation and why it's so critical for telcos. So built on top of the business case that Niall's been explaining. So first of all, network automation is a software platform that sits in between the business systems that you see on the left. Those are the systems responsible for the catalog of products that you can buy from your provider, their billing systems, the care systems and so on, and then the infrastructure on the right. Niall has already mentioned the networks are moving towards being a lot more software based. So infrastructure these days is not just the boxes and cables that we used to associate the network with, but they also involve edge, cloud and IT platforms, they're becoming a lot more complex. Sitting in the middle, we call network automation. There are four fundamental roles of any network automation platform: infrastructure rollout, rolling out everything you see on the right, it uses workflows to ensure the process runs smoothly, and it also uses inventory systems to record everything that is deployed in the field. And we have service delivery. Once you've rolled out your network, we're going to want to sell services to your customers. Service delivery is responsible for your mobile connections, your VPN, broadband connections. And it also uses workflows and inventory to support itself. Then we have assurance. Infrastructure services, they need to run and run correctly. They need to be fixed, if needed and ideally before it touches the customer. And then operations. Again, Infrastructure Services have a life cycle they need to be upgraded, replaced, repair and sometimes move and that's the goal of operations. So network automation really is a critical part that sits at the heart of the telco operations and is absolutely vital to make sure things are delivered reliably on time in a cost-effective manner. We move to the next slide, please. Another two reasons why automation is so critical is because the scale and complexity of telecom network these days is so great that it cannot be handled manually. The diagram you have on the screen is a very high-level simplified schematic view of a network. Your average midsize telco will have -- over 100,000 of those boxes that you see in the diagram that represent hardware and software. They will have over 10 million connections. The solid lines between the boxes. Over 50 million customer services, the way dotted lines. And all of that spread geographically across several data centers, cloud storage locations with different IT, security, network equipment vendors and maintaining the new technologies like 5G, and the legacy ones, the 2G, 3G, 4G, because all of them coexist for many years. Now think back to the role that we just described of deploying and maintaining all of these, would you do this with by hand with very low automation. If you want a healthy network. No. So let's move to the next slide, please. Now the environment in which automation operates is really changing very rapidly, and Niall touched on that several times. First of all, it's becoming software-driven. The hardware and software are getting the capital disintegrated. Now the hardware doesn't affect as much because we're not a hardware vendor that, what happens to the software is very close to us and has this severe impact on automation. The software is beginning to get broken down into smaller atomic pieces of functionality that no longer have to be placed in the same place, but they can be located in different edge locations or cloud locations wherever it makes more sense for them to support their network. They can also be dynamically moved. They can be dynamically sized. They can be scaled up if more capacity is needed or scaled down if it doesn't, in order to conserve resources and save money. These changes have brought many more vendors and technologies into the equation is in general, is all very positive movements from the industry that make the networks better, but it adds complexity. Before we move on, it's very important to mention here, satellite, especially since we -- our next speaker in this space, satellite is becoming a very prominent, very full from technology. It has had a lot of improvements in the last few years and is beginning to offer really competitive broadband services, especially in rural settings. And it's beginning to get combined with cellular technologies in order to broaden the capacity of mobile networks. But is also evolved quite significantly is the customer expectations. As consumers, we want to choose what we want. I want 10-gig. I were roaming, but I want roaming, and I want to pay for it for 2 days this month but not next month. And we want to be very flexible on what we did and when we get it. We also won services in minutes. We've all grown used to timeloading applications from the App Store in our phones and getting things in seconds, and we're no longer prepared to wait for weeks or months for any service. And reliability is getting incredibly critical, especially in industrial settings, where connectivity failures can cause equipment failures can cost customer money. So in general, telcos and telco networks are looking at a lot more complexity, much more agile and dynamic. And they have had to adapt the automation platform in order to cope with these and they have. Could we move to the next slide, please. The best way to explain how they've evolved is to compare it to how a taxi ride experience has evolved. If you think about what it used to be like when you used to jump on the taxi, either the taxi driver knew where he was going or you would go, go straight ahead, turn left. If there was an accident or a traffic jam, you wouldn't see it until you were on top of it. Think about how that has changed these days. Even before you jump in the taxi, you can order your taxi in advance, you will know how much is it going to cost, more or less, how long does it going to take, then we jump on the taxi and the taxi has a satellite application system that looks at the current conditions of the road and gives you the best possible route for that moment in time. And even when you are on road, it's notify or say a traffic jam, it will redirect you in a different way. It's much more customer and service-centric and that is exactly how automation platforms have evolved. They are no longer procedural driven for delivering a service, do this, do that, do that. In the world of software-defined dynamic functions moving around and being real time is impossible. Automation has become what is called intent-driven. Essentially, you tell the system what you want to do, not how and the system works out the best way of doing it. Experts are very important. People are very important, but they're beginning to need the help of intelligence to cope with the real-time nature of the new networks. It's no longer acceptable for things to break down and customers to be waiting for days for something to be fixed because it's been done manually. Automation platforms are helping in this aspect by becoming -- by providing predictive capabilities, observability capabilities that allows for things to be fixed and corrected even before the customer is affected. So in general, much more customers, service-centric and much more intelligent. If we move to the next slide, please. I think the best way to try and visualize these for -- in a telco context is to look at an example. Let's imagine the business system update on the top right sends a request to deliver a premium gaming service for a customer at a given address. The automation system is going to receive this as a business intent. This is what I want to do, and is going to decide what to do based on the parameters it knows that, that service needs. For example, this type of service needs less than 10 millisecond of latency, and also in the current conditions of the network and what capabilities it has, like what is my closest edge location, if any. Based on all of that, it's going to send the activation commands down to the network, the actions. The service is going to get provisioned and then the system is going to start receiving feedback and listening to what is happening on the network observing, detecting. Observability is key. For example, the latency of service to going up to 20 milliseconds. We're going to starting to a diagnose mode, but we try to work our bodies from. For example, the services that we utilize. This information is going to be passed back into the loop, back to the beginning. So the system can work out what to do now and what to change in the changed conditions in order to fix the problem, ideally before the customer notices. At the center of all of this, we've got data. Without data, none of these works, data and artificial intelligence from Machine Learning is absolutely key in making sure these systems work. Let's move to the next slide, please. Now this is our platform, Amdoc's intelligent networking suite is based on designing the same principles we've just explained in the previous two slides. We have four main components. On the left, we've got the design which is used to specify the parameters of the services, like we mentioned, the gaming service needs 10 milliseconds of latency. That design, the models of that design are send to the run time. And in the run time when a business intent is received from the BSS systems, the role of orchestration is to look at the models, to look at inventory that has a record of what is available in the network, to look at assurance in case there are any alarms coming from the network and then decide how to best fulfill that service. The service will be recorded in inventory for the future and then when everything is activated in the network, the feedback will start and assurance that will start monitoring everything in a closed-loop manner, okay? Now we have a little video that shows those components working in the context of 5G services. Could we please play the video? [Presentation]

So if we move to the next slide, please. Thank you. So we saw a couple of slides back, a small example in the context of the macro network and a consumer service. Let's look here an example in the enterprise context. Automation in the enterprise world is becoming really important because the services are getting much more complex. This example is a project we deliver in a farm, in the north of the states. Where we had to deliver three main services that you see on the left. So we had a drone providing soil humidity and crop disease images and analysis. Then we have an AR/VR solution for remote equipment maintenance and video analytics, security and surveillance solution. Towards the right, we also have to provide a private network and an edge because there was no cellular coverage in the farm. This private network was connected to a public network provided by T-Mobile and then connected all the way to the Internet and the Cloud where several of the applications and platforms, including Amdocs Intelligent Network in Suite were deployed. So the rollout of our platform in this scenario was, first of all, to roll out all of that infrastructure in the private network around the core of the network functions to roll out the edge, then roll out the three applications on the left, on top of that infrastructure and then deliver or fulfill the services on top of the infrastructure and the applications. And make sure that everything was stitched entwined. Because if you think about it, if you have a drone capturing images. Those images are going to go to the private network and then to the public network by T-Mobile and all the way to the cloud, that has to somehow be connected. That's what we do. And finally, we also did the operations. We have a little animated -- a few animated slides that show the drone example and how it works. So I'm hoping it gives you a sense of why automation is needed. We started flying over the fields at 7:01 like [you sense] there. The drone has started scanning images, and at 7:02, those images were sent to the edge server in the farm. It's too much data to send all of it to the cloud. So we send it to the edge first to compress it and stitch it because otherwise, it's too expensive. At 7:03, once all of the data had been compressed, we sent it via the public network to the Microsoft Azure Cloud for analysis. At 7:04, one of the partner applications took all of the data up in the cloud, process it and derive all of the intelligent analysis. And at 7:05, the data was sent back from the cloud to the farmer, and you've got there a couple of those are real screenshots from the application provided insights on crop density and humidity. So thinking back of what I said of automation, all of that process of capturing the images and sending it through all of the part needs to be stitched, that has to be automated. And not to mention the delivery of all the infrastructure before. And what that does, if we look at the results is -- apart from the benefits at the top, it reduced the process that used to take days to 4 minutes, as we've just seen. But for the farmer, which is the most important here, it increased the yields of the crops. It reduced water consumption by 50%, and it gave the farmer better visibility of where to [ water ] for future years. So that was the end of my presentation. Thank you very much, and I hope it's been useful. Thank you.

Ana, Niall, thank you very much for that. Really good overviews. And I think you did a great job in outlining exactly the market need for network automation, and of course, the way in which Amdocs is very well positioned to support our customers in this area. Now speaking of our customers, it's my great pleasure to introduce Gint Atkinson, VP of Network Strategy and architecture to SEC -- SES, sorry. SES is one of the world's leading satellite owners, operating more than 70 geostationary and medium-orbit satellites and delivering video and data connectivity worldwide. So Gint, Welcome, and thank you very much for being here today.

Thanks for having me join.

Yes. No, it's a great pleasure. Now, just to begin, maybe just a little bit about your background, starting [with us]. With your days back at T-Mobile U.S. and all the way through to your current role with SES, just to give a little flavor for your experiences and the career path you've taken?

Yes. So I would say my deepest first experience with Amdocs was really a T-Mobile USA, where we were launching among the first real-time dynamic usage-based data services for wireless packages. So this is all the way back in the early 2000s. And I was doing that after we launched it -- launched EasyWeb for [ KDDI AU ] in Japan. And the striking thing back then, this was all the way back in the days of Samsung. And at T-Mobile, we were busy launching hundreds of new features and dozens of new service packages a month. This is long before we had the well-articulated discipline of agile methodology and so forth. And Amdocs was extremely fast and extremely disciplined. When we set our goals and our objectives with Amdocs and agreed to them made a project plan, Amdocs made sure we got there. So working on a variety of other accounts, going from the early 2000s and back to a lot of my work in APAC, where I was working with customers like Globe and PLDT, when I was CTO at Colt Asia and also working with Ciena. So I was constantly coming across working with Amdocs on integration projects. And I always appreciated the extremely serious discipline that Amdocs brought to the table. And this is a good 15 years ago. Now the Amdocs today that I encountered with -- at SES, it was the Amdocs that was really innovating and acquiring the most innovative software and technology companies in telco and with all the work that Amdocs was doing on e-com with AT&T, quite simply put, the network automation that I wanted was [ on ] app in the Azure cloud with a complete solution for standardized connectivity services like SD-WAN and [ math-based ] Ethernet connectivity services. And then further in the background would be everything that we do with content because as a satellite provider, we're delivering content all over the world, more than 3 million homes and connectivity all over the world dominantly to remote enterprises. So this is really how I wrapped it all together and -- but the recent experience was that Amdocs is this extremely agile and innovative company, but at the same time, incredibly disciplined as I had experienced more than 15 years ago. And that was something I was hoping to capture so that we could take bigger risks, we could take larger jumps and we would have a vastly greater chance of succeeding by having Amdocs take us there.

And of course, SES did select Amdocs for a lighthouse project to deploy our end-to-end service orchestration solution back in 2020, we can talk about that a little bit in a moment. But when you -- when we think about the historical orchestration challenges that satellite providers such as SES and others are faced. Can you talk us through a little bit about those difficulties? And why is it a new sort of approach to orchestration is needed.

Yes. So where the satellite RAM infrastructure industry is at today. When you look at [ SAP RAN ] gear, it's exactly where terrestrial ran gear was 15, 20 years ago. Everything is completely proprietary. So today, when you're a SAP RAN engineer, and you have to figure out how to operationalize things. You don't talk about the service chain. You talk about the RF chain, and that's the antenna, the buck, the amplifiers, the filters and they all have APIs and they're all proprietary. There are some standards there, but you're literally dealing with things at an RF component level and an operator, a satellite operator or a teleport operator needs to be a systems integrator and tie all of those components together. So what we really wanted to do, what SES wanted to do was to focus on pulling our RAN vendors together, doing our own internal innovations on developing controllers for that RF chain that would be multi-vendor neutral and stitch that controller together underneath a resource and a service orchestrator. So where we ended up was creating a relatively clean SAP RAN domain and then the rest of our service domains being the transport domain and the cloud domain. Those were all plug-and-play. Amdocs was ready to go. Ethernet service over those network domains were completely standardized. So with Amdocs bringing all of that to the solution -- to the solution table, we really just needed how to figure out how to integrate all of those RF chain components, all of the components in the SAP RAN, put it under a controller and do some basic domain orchestration and then we just plug in straight into service orchestration and resource orchestration from Amdocs. We of course, we also added on top of service and resource orchestration, we brought in both service and resource inventory, which was necessary and wrap that up with service assurance and closed-loop control. So now we can do advanced things, for example, we see a weather pattern, a cyclone approaching a gateway, a serving gateway. Now what we can do is reconfigure the resources, give more power to the antennas, give more power to the right resources to support the most critical components that we're serving at that gateway so that we can get through that weather outage and all of this is exactly as we saw just in that diagram, the end-to-end all the way from service design, all the way on through to service assurance and closed loop control. So the exciting thing we did was to integrate not only the network resources, but we're also integrating in the weather data that we have and other forms of threat to the resources that would impact services.

And if you don't mind -- Sorry, just to say -- the reverse is true as well. I think when in times of good weather, you can tune the power and the network and the resources to be optimized for those more favorable conditions.

Yes. This is -- I used the example where we're just strictly speaking about ground resources, but the scenario you brought in ends up looking at the full RF chain going all the way up into the sky and to the satellites. And the satellites, for example, our Empower satellites, what I think we have 5,000 beams per satellite, for example. We're able to deliver multi-gigabit services. So exactly as you pointed out, with 8-plus satellites in orbit for that constellation, we have thousands and thousands of beams served by 8 satellites, solar panels going around in orbit, burning up the battery, recharging the battery and delivering services. So that equation gets extremely complex just at the satellite level, but then you have to optimize it all the way down to the ground level as well because we're typically picking up a remote enterprise, mining, energy or a remote cell site, and we're frequently direct connecting them one hop away into a direct connect to a cloud provider like Azure. Of course, on the Azure side, where we have the gateways co-located with the cloud, there's massive efficiencies available there. There is progressively more and more infrastructure that we can virtualize. So you can expect that the SAP RAN architecture is migrating towards that cloud ran style architecture where baseband processing is going on to an FPGA farm and so forth. All of that is coming. We don't have it today. We're working on it with our vendors, with cloud providers, especially with Microsoft. And we're going to see a lot more innovations that we're going to push from terrestrial RAN into the SAP RAN. And then we're also hoping to bring satellite brand innovations into the terrestrial world as well. But definitely, resource planning, capacity planning, RF optimization, large-scale spectrum planning, satellite brings new scenarios to it. And we, of course, need to tie in the most advanced spectrum planning systems and optimization systems to make that. end to end allocation of resources and spectrum.

Yes. So when we think about the operational benefits of this Gint, in terms of thousands of processes which are automated, I presume they reduced the fulfillment to order times and so on. Can you hang some meat around that for us so that we can invest to...

I'm going to share a couple of details that maybe on the terrestrial side, everyone keeps on overlooking I think drive buys. We've been doing drive buys with antennas on cars and driving all over the place for a long time and then cell sites got redesigned to start sharing that RF data that everyone's antennas were catching and start capturing it into a database and sharing it so that people could see what the trim was in a specific location and look at the interference over time. And this is something that we haven't done as much with satellite players, and we definitely don't -- we haven't opened up as much data and shared data between all of the spectrum users the way that the terrestrial mobile industry has. But -- the net result is that when you get an order and you want to say, I need to provide 10-gig service at an industry -- a remote industrial location. Do we have the capacity available? Are there going to be any interference issues and net-net, part of that whole -- part of the order and rather the quote, is to do a detailed link budget. So in the satellite business, you'll see a sales engineer working with a variety of tools to produce 2 or 3 pages of very detailed link budgeting information to support that quote. If I want the 10-gig terrestrial connection between the 2 locations, we can go on the web and specify a data center address a, data center address b, and we'll have a quote for a 10-gig Ethernet circuit within minutes. So the gap is huge. And one more thing, building out a 1-gig endpoint or 500-meg endpoint is a little bit more analogous to building out a small cell site. You may need to pour concrete, you may have big antennas, you may have big power requirements. So you've got to sort that out, your installation may take days and so forth. But it's pretty common in the satellite industry that it is months, many months from order through fulfillment. So to be able to compress that just to the hard physical labor, pouring the concrete and installing the antennas, there's a lot of other stuff that adds days and weeks, that's the key thing. And like in our case, we had a good thousand processes that involve engineers, installers, planners, and a massive amount of that could be close to 100% automated or even eliminated.

Amazing. It's fascinating. Thank you. So just to wrap up a little bit here because we are 10 minutes away from the top of the hour. What is your version of network automation for SES going forward? What is the journey look like over the next several years?

So the service orchestration now we've planned out using standardized Metro Ethernet Forum services, not only point-to-point, but all the way to multi-point, E-LAN and E3 services, along with SD-WAN services, in particular, the maritime industry really has great use for SD-WAN-based and terminated services. So this is something we've built up. We started off with E-LAN and we got massive reuse across the different types of Ethernet services we were building, and that all fed into SD-WAN. And all of this is kind of wrapping up, getting closer and closer towards network-as-a-service. So defining and onboarding these services is really fast and easy for us. We're able to do service composition and use underlying component services to define new services and to create complex service packages that includes support, installation services, getting the kit out to the site and so forth. And really, where we're going to go -- the next explosion of complexity is beyond connectivity services, and that's onto our video business. And this is where I'm really interested in. Amdocs' deep experience in video services and everything that you have to offer there, because video services are complex. You may need to do ad injection. You may need to do ingestion and upload a video and then play out, it's extraordinarily complex. So I'm really looking forward to how we can design these services in a modular fashion, compose them and launch them very, very quickly, test launch revenue. So this is really what Amdocs has enabled us to go to a much more agile model, maximizing reuse and modularity so that we can do much more dynamic service composition. And then there are very much smaller pieces of service components to innovate from scratch.

Gint, just to, I mean, look, I'm very pleased with the journey you've been on, I think your leadership, and I think the SES organizations embracing of the automation journey you brought them on has been phenomenal. To put it in context for maybe the wider audience, the automation of these kind of processes, which were quite manual. I mean -- would you say you moved dozens of services, tens of services, hundreds of services just over the course of maybe the last 3, 4 years.

Yes, dozens of services, and we're launching on the order of dozens new services that are vastly more complex, have -- where before, we may be forced into doing a bespoke service quote. Now what we're talking about is twisting a dozen knobs and dials on service characteristics. So what we can do is we can characterize services with 10, 15, 20 attributes and then some packaged add-ons, so we're getting increasingly modular instead of bespoke where everything starts from scratch and there's the tendency to take the smallest little pieces and put them together in different ways here. We have many more building blocks that get assembled. We're going -- I feel we're going much more into an assemble-to-order model with very, very trustworthy, well-defined service components.

Gint, thank you so much for your insights. I mean I think it's a real treat for investors to be able to get the customers' perspective because it's not always easy to do and -- really appreciate you being here today. Some fantastic data points, and we look forward to connecting with you again in the future. So thank you very much for your time.

Many thanks for having me.

All right. So we've got just on 5 minutes left. Ana, Niall, so I'm going to put you on the spot here a little bit and ask you a couple of questions in a Q&A session. Okay. So -- maybe if we can just go back to something that you touched on in your presentation and maybe just expand on it a little bit, but it's around the idea of competition and the existing landscape and so on. Why is it exactly that Amdocs believes that we have a potential competitive advantage here within the domain of network automation versus, say, the traditional players who are obviously coming from the hardware side. What is it that you think puts us in that spot to be able to really harness this opportunity?

Yes. I guess you look at the traditional landscape where the network was largely serviced by some of the bigger equipment providers, the Ericsson, Sunakis, Huawei, if you will. What they've done historically is provided network elements, physical boxes may be moving towards more virtualized kind of network functions, where Amdocs have a competitive advantage is probably in a couple of areas, as I see it. One is a lot of the business outcomes, the objective of making some of these changes. It is heavily driven ultimately by monetization models and by IT system type characteristics. Networks are becoming software platforms, not physical chains of individual assets. And I think Ana touched on some of the complexity around that. So the Amdocs kind of end-to-end visibility view of money all the way through to a network transaction, which we built over many years is something that the equipment providers don't always have. So we have an insight around the business process is not generally available, where some of the equipment guys may have more engineering focus. The ability to look at these systems and realize that it's more than just the boxes that there are, as Gint talked about, business processes, there's people, training processes, there's a whole kind of a suite of different skills required to take a marketing vision and translate that into an end-to-end kind of flow of people and processes and the technology to make it work. And finally, I guess it's around the agnosticism. I mean, again, Gint talked a little bit about the -- some proprietary technology and networks, some standard technology and networks. There are very few networks on the planet that don't have multiple big equipment providers. And when you're trying to orchestrate services across the full end and automate those you need an agnosticism, you need an overlay, which can plug into any type of network profile and modernize those and orchestrate those capabilities without being proprietary. And I think for all of those reasons, I think it's a catalyst to be a change agent to be a business partner. At a time of unprecedented change for the networks as they embrace this new automation technology. That's really for the first time ever given Amdocs this enormous play.

Okay. And then maybe a double-barrel question just to wrap things up here because we're almost out of time. When you look at the pipeline and the way the pipeline of opportunity within network automation and how that's developing and then you look at the massive amount of work that needs to be done over the next several years. Can you just talk about the longevity of the cycle here and what it's going to take to actually get the industry towards more of a fully network automated sort of round.

Yes. I mean I think the short answer to the question is it's multiyear. It's probably a 10- or 12-year transformation cycle in most operators not because people are moving slowly, but you've got existing technology and new technology, which keeps getting innovative. You've got a multitude of new business opportunities and efficiencies that can be garnered. So you've got a lot of spinning plates to keep in the air and nucleates being added all the time. We're at the very really, the very early stages of this, all of our research, all of our engagement is pointing to a kind of an early stage is being implemented in multiple operators. I'm very optimistic that this is going to be something that we can play a significant part for the next decade, possibly beyond because new changes will come from there but it's very much a transformational leap that will be done in measured stages with gates that have to be gone through for success. So it's not a, write a huge check and everything will be fine trust us. It's going to be -- in a network domain, it's a much more staggered, measured process of transformation over time.

Niall, thank you very much. Unfortunately, we are out of time, but I would like to thank just Niall and Ana for their fantastic presentations today. And particularly, again, a big thank you to Gint for taking time out of his own busy schedule to -- with the SES to come and be with us today. I know we didn't get too many questions, but please reach out to Jill and myself in the Investor Relations department if you need anything further. If you want to continue the conversation. And with that, have a great rest of the day.

Thank you.