Thank you. Good morning, everybody. Welcome to Stifel's Cross Sector Insight Conference 2024. I wanna welcome you all for being here. My name is Shlomo Rosenbaum. I'm the Business Services Analyst here at Stifel, and I wanna welcome Shuky Sheffer, who is-
Thank you.
... the CEO of Amdocs, and open it up. I just want to start, Shuky, with just a very, very basic, give us the one-minute elevator pitch of what Amdocs does, and then we're gonna jump into Q&A.
One minute? This will be short. Amdocs is doing all-
You can do two minutes if you really need to.
Okay. Amdocs is doing all the platform and systems for the what we call the service providers, you know, the telcos of the world, the AT&Ts, the T-Mobiles of the world. And if you need to think about it, from the touchpoint that you engage with the service provider, could be a website, mobile application, a retail store, then to all the ordering systems, how you order the service. I mean, it could be simple service, it could be complex service, if you want to bundle fiber, broadband, and mobile. So all the ordering systems. Then, after we are doing all this ordering system, then if you go south, then obviously, all the system that support all the financials, all, you know, billing, general ledger, all the financial system.
Then if you go deeper, then there is all the provisioning system. Eventually, if you buy a certain broadband service with a certain speed, someone needs to provision this service to the network. And then all the system that actually is actually activating and doing the orchestration in the network. So from the channels all the way to the network. I think that we are a very unique company from the perspective that there are software companies, Salesforce, ServiceNow, and there is a service, there is services companies like Accenture of the world. We are doing something very unique because what we, we are what we say, product-led services company, so we develop the products. By the way, we are working, for the most part, only the telco market, in the service provider market.
So we are developing all the software to support this company, and we are doing all the services. So we are implementing our products, and we are operating our products, which create a very, very unique accountability model. Usually, when you buy the software one company and the system integrator implementing it, if something goes wrong, the system integrator will say, "The product doesn't work." The product say, "The system integrator doesn't know to implement this." So it's a very unique business model. We are market leader in our domain. Roughly close to 3 billion people are touching Amdocs on a daily basis. It will be very difficult for you to go to, in the U.S., Europe, APAC, Latin America, without open your phone, without touching Amdocs.
As I said, it's a very unique business model. We are focusing on the telco industry, and I think that we are by far the market leader in this domain.
Okay, great. And can you touch a little bit on the two largest clients that you have? You do with T-Mobile, AT&T. You know, can you talk about how those contracts have evolved over time, the expansions of those contracts? And, you know, you really had recently had an expansion of AT&T contract, and that involved, I think, some of the unique capabilities that you did with the recent acquisition. Maybe you could talk about, you know, how you're operating there and what you're able to do across the communications industry that others cannot do.
Okay. If you talk about specifically on AT&T and T-Mobile, you're talking about tens and tens of years of partnership. I mean, we actually started with AT&T when it was in the nineties. It was SBC. We helped actually AT&T to when they start, you know, this acquisition trend, and they bought many, many companies until they become AT&T. By the way, we can talk a little bit about that later, but we are usually in a very good position in a situation of merger and acquisition. In the conversation, we talk about this. So we are supporting AT&T for probably 40 years. It's our largest customer, and we obviously, we evolve within AT&T.
We're very connected, and this is, I think, also related to Amdocs, how we evolve our product and services. So if you look about probably in the nineties, we were a billing company, then we added CRM, then we had a network and ordering, many, many. So if you look at the portfolio of Amdocs today, pretty much covers end-to-end. Back to the definition of Amdocs, is usually in an RFP situation, it will be Amdocs that have the whole suite of products pre-integrated. And usually, we will, we will compete with a consortium of a system integrator that try to stitch different products together because we have the whole, the whole portfolio of the products.
And so we grew up in AT&T, so obviously today, all the, all the consumer, mobile, consumer in general in AT&T is running on Amdocs system, both mobile, broadband, et cetera. And we have, if you—it's not just the, the AT&T brand, it's the Cricket brand, it's AT&T Mexico. We are doing a lot of data activities for AT&T. So if you look today, we are working with many buying towers in AT&T. By the way, many, many of our activities are non-Amdocs system, but they touch our system. So even if it could be a different legacy system or other system that are touching our platform or our domain, we are running them also. Another something that we evolve in AT&T, definitely many, many, many other customer, is our managed services business.
If you look at the Amdocs revenue today, roughly 60% of our revenue come from managed services, which is more like long-term agreements. And we have roughly 80% visibility when we go into start the year, you know, our financial year. So we are doing a lot of managed services for AT&T. On the deal that you touched, and this is when we talk, I guess, more on the strategic growth domain of Amdocs, definitely the journey to the cloud is one of them. By the way, just to give you, today, probably, if you look at all Amdocs customers, and we are the market leader, less than 10% of the workloads move to the cloud.
So the industry as a whole is in the process of moving, but they are not early adopter, let's put it this way. We are working with all our customer to move to the cloud, but this takes time. The deal that we've signed with AT&T have two components. One of them is we extended all our agreements that we have with AT&T until 2029, consumer agreements and other agreements, which is... I'm not-- I don't think that someone in AT&T wants to replace us, but it's good always, you know, to extend your agreement in the consumer domain and all the other domains.
And the other thing is that when we took giving our position ourselves to make sure that we help the industry to move to the cloud, so there are certain elements in moving to the cloud. Some of them is our platforms, or our platform today, obviously, cloud native, so we allow our customer to go to the cloud. But at the same time, we have the capability, with our partnership with Microsoft, AWS, to move other application to the cloud, which are in our domain, but not necessarily Amdocs application, also to move to the cloud. And another thing that we found out, that there is a certain, I would say, domain, which is a mainframe.
So many of our customers, if you guys have really old mainframe application, and in last November, we've done an acquisition of company called Astadia. We're looking how to automate this. So we looked around, we found out there's the best tools in the market to of converting 40, 50 years old COBOL application to Java running on the cloud. So we acquired this company, and actually we're using this technology, and we signed a 5-year agreement with AT&T, which actually we're taking all the AT&T mainframe application, and during the next 5 years, we are going to move all of them to the cloud using this company. So this is an example of how we are helping our customer through the cloud.
Generally, if you look at the cloud, today, it represent a little bit more than 20% of our business, and it's going double digit, so this is going very well for us. You mentioned T-Mobile. T-Mobile is a great customer. It's a great success story for Amdocs. If you look today in T-Mobile, this is actually three Amdocs customer. It used to be, you know, the T-Mobile brand, and then, which we are doing many, so many modernization activities to T-Mobile. It's also MetroPCS, which used to be a customer of Amdocs, which was acquired by T-Mobile, and Sprint, which was also a large customer of Amdocs.
We helped T-Mobile with all this, after the merger to do this. I would say great. I think it's a great success story in bringing Sprint to the T-Mobile system, et cetera. And now it looks to be that there're going to be additional opportunity for us, given that it was announced, I think, a week ago, that that T-Mobile is planning to acquire UScellular, which is another Amdocs customer. So it give us a lot of opportunity to support them in doing this. Usually, it come with a very large consolidation projects. And I think that, as I said, we are a strategic partner with T-Mobile. We were supporting them on the back end with all the Un-carrier, you know, amazing story and supporting them today, too.
Okay, great. Thank you very much. Can you talk a little bit about how the macro environment impacts Amdocs? And, you know, you talked about you have managed services as a large component of the business. You're mission-critical, you know, so obviously, there's a lot of the business is just repeatable. But there is a part that we, you know, you talked about last year that the, in a macro downturn, you know, clients are a little more hesitant to invest in their legacy applications. Maybe you could just talk a little bit about how that works and how we should think about that going forward.
So you're very right that what we do for our customer, it's a mission-critical system. This is not discretionary. If God forbid for some reason Amdocs platform are not working, everything is shutting down... this is why we don't sleep at night. You know, usually when other people enjoy the we say Amdocs the holiday season, you know, like Black Friday, this is the largest I would say operational activity for us. So back to the So you said you're right we're running mission-critical systems, and this is not something you can turn on and off.
This is so discretionary spending on this, it's something that you cannot say, "Okay, I'm stopping development." The phenomena that we've seen probably in the last, we started to see it probably. Remember, our fiscal year is October to September. So we started to see it in mid 2020, our 2023 fiscal year, is that because of all the macro environment, and I will say, what does it mean, macro environment for us? We saw customers starting to prioritize their investment. So if you look at the business, Amdocs business, the 60% managed services is pretty much solid and untouched. When we talk about discretionary spending, they put the brakes a little bit in what we call the legacy investment.
By the way, legacy, if you think in most of our customer, we have an environment that we are running for them, in managed services, the current platform. By the way, the current platform are the same platform they are competing with. I mean, it's not that, this is a... If they want to go now with some new device, if AT&T wants to go with a new device financing option for Apple launch in September, this is all these changes and implementations coming on what we call legacy or current system. While at the same time, we build for them the next generation platform, which is obviously our 5G related, on the cloud and everything, that generative AI capabilities. We can talk about generative AI in a second. So this is happening in parallel.
So when the macro, obviously, they put some pressure on everyone, including our customers, and they put the brakes on CapEx. By the way, we are less impacted by CapEx. CapEx is more mainly, mostly network. We are more OpEx. So when they look, you know, to cut costs, you know, to make sure that they meet their, you know, free cash flow and all the parameters, the financial parameters, then they say, "Okay, in the IT domain, we don't want to touch the future, because all this platform that we build right now, this is going to be our future." So they put little brakes on the current platform. By the way, it can never go to zero because this is how they compete today.
I mean, but they can, the way they do it, by the way, say, if before, they wanted, 9 months ROI, now the same one, 6 months ROI, so immediately some of the changes are falling. We've seen, Sorry, when we got it to 2024, we said we see roughly 3% headwind, and this is pretty much stabilized, so we don't see it happening again, next year. But this is where they started to do some prioritization. At the same time, they are doubling down on investment in the new platforms, as I said, cloud and other generative AI capabilities and et cetera. So this is the environment that we see right now in the market. We don't see it. It's pretty stable.
We don't see it, you know, getting worse, but we don't. On the other hand, we, like everyone else, are waiting for things to ease up a little bit when the macro will change.
So help me out a little bit, just when I think about this. So we had this, and you had tough comps going on that'll end up happening through, say, I guess, the summer, right? You know, where you started to notice that in the middle of last year, and it kind of flowed through the June quarter. Once you get past that, and we're at a steady state, do you end up with a naturally better growth rate because you've already comped on that? Or is it that things are just slow, they're slow, and you kind of settle in at a kind of a little bit of slow?
What we said, I think, when we gave the guidance for the year, is that when the customer slowed down investment in legacy, the impact on us is almost immediate. Because we are doing this, and they say, "We want to stop it. We've done this quarter this, and..." So this is why the headwind is almost... Now, the revenue which replaces this is all the new projects that we are winning, and we are seeing that we had loss in our domain. So all the new projects that we are winning. By the way, remember, we're talking about the part which is not managed services.
The new project we are winning, it takes us some time to ramp up, to cover for this, headwind, which is almost immediate, and this is why we said that we were going to see relatively flattish, half one, which we finish, and we are going to see, starting to see acceleration revenue in Q3 and Q4. This is when... Now, the headwind is not happening again, it's already in our numbers, so it's not... So now, actually, it's how we are going to ramping up all the, the project award that we are winning and starting to ramp up revenue and starting to... So we did say half one is going to be flattish, half two, we are going to see some acceleration in revenue.
As the new business coming through that, I understand the legacy kind of took a step down. Is the new business coming through at the pace that you were expecting, or is there any change in the new business as well?
... Almost, probably not exactly, you know, when you try to plan.
Yeah.
I can give you an example that, the AT&T deal, which is extremely important and strategic in nature because, you know, these are larger customer, give us visibility, and it give us a very, very significant incremental revenue in domain that what is new for us. This is not in our backlog, and we thought that we planned that, it will close in, in Q2. Eventually, it was closed in Q3, and so it's not even in our backlog with me now. So, things are a bit slower. We don't lose anything, but, the trajectory is the right one, but maybe a little bit, moderate, comparing that, that will...
Another thing that I don't think I spend enough time is how it's also connected to the business model of Amdocs is generative AI.
Yeah.
Okay.
This'll be my next lead in. Maybe let me ask that question. We'll take it in that direction.
Okay.
I was just gonna ask you, you know, where have you made your investments in generative AI? And talk about the part that's, you know, revenue generating or client-facing, and what you're doing both internally and, you know, just give us a broad base and maybe a couple of examples.
Okay. We started to be very active in generative AI actually more than a year ago. We have several strong partnerships. Among them is Microsoft and very strong partnership with NVIDIA. And I think the best way is to explain Amdocs' role in generative AI is with couple of examples. So, let's talk about product catalog. Our product catalog is the most, I would say, by far, the most advanced catalog today. Catalog is where you define all the offering of the company. This is the heart of all the system of the company. So today, it's a very advanced catalog.
We can come in and say, "I want to build this offering." You drag and drop different, trying to guesstimate what will be the uptake of the new offer, et cetera. So now, in the new catalog of Amdocs, actually, you talk, I mean, you say, "I want to build a new offering for Dallas area, for millennial in this age, that are likely the students, that are likely to change their, upgrade their phone," and this and this. Immediately, it will create you, like, these four different options to pick from. So it's this within this every almost every product of Amdocs is coming with a Copilot. This is more within our products.
The next example is what we see, and I know, and I know that it's not that sexy, but, we are doing all the plumbing that you need in order to to leverage generative AI. A good example is, you know, when you look about call center efficiency, so, we, we've done with one of our customer, we took 50,000 calls from the call center. We ran them through speech-to-text, and we found out, not that we were surprised, that the most common and complex question is, "Why my bill this month is higher than last month?" Now, for even a very experienced call center rep, it might take maybe 7 or 10 minutes to answer because you need to look... It can come from 10 different reason.
It can come because a month before, you were pissed off, and you called the call center, and they gave you $20 billing adjustment, and you run out of a promotion. Many, many things. So the way we are, so the customer, they say, "We can do it ourself." They took the two PDFs, send them to OpenAI. It takes, like, 40 seconds to get a reply, which was very expensive because, you know, OpenAI charging by text, and the accuracy was 60%. Now, because we understand the fundamental of this question, we actually, with our platform, we can bring all the relevant information in real time. For example, we check, did you change offer? Did you get a billing adjustment? Did you have a promotion that ran out?
We are bringing all the information, by the way, using, in this way, using mainly NVIDIA and the capabilities, and eventually, when... By the way, we publish a PR with NVIDIA. When we are sending this to OpenAI, which is the best, we, our accuracy is in high 90s. We reduce the latency to almost anything, and the cost is very affordable because we are probably sending 90% less text that you need to, because, you know, when you send OpenAI everything, so, and, so when we are building this, we can create huge efficiencies in the call center. Now, when you do this, it has to be mission-critical system because if someone, a customer, take a decision based on this copilot to reduce 10%-20% of his call center employees, this have to work because if suddenly-
Right.
So, how you're building all the plumbing, taking the data, and making it a mission-critical system, Copilot, that can change the call center fundamentally. This is something that we do, and we have, and right now, you ask about. This is, first of all, it's something that we have to do, as the market leader in this domain. Second, we see also revenue growth opportunity. We are running, as we speak, you know, probably 10 different proof of concept with customer, and we got already couple of awards. This is a major data play because the whole idea is how you bring the data at the right time to answer the question in the most effective way. So, this is something that we are very bullish about.
Another aspect of generative AI is related to our internal... and we are now in the process of taking this. If you look about our margin expansion, while the gross rate slowed down in fiscal 2024, the margin expansion was huge. This is why, because we're starting to see how we can implement this technology in our software development life cycle, in our operation. Now, I think it's important, because at the end of the day, everyone wants to say, who is the winner of losers in generative AI? And if you look at in our domain, we have a very different relationship in the managed services comparing to our competitors. Our relationship are outcome-based, so when we have a managed service agreement, it's not a...
They're not buying from us. It's not a rate card relationship. They're not buying from us 100 people in certain rate card. They are buying outcome. The system has to be up. It has to be all the time in line with down R&D. I mean, it's mainly operational parameter, which are outcome-based. So in a way, obviously, we always need to share, you know, efficiency with our customer, but I think which we are much more protected, giving that all our agreements are outcome-based, comparing the traditional IT services company, which most of their agreements are mainly, you know, rate card-based, and... So it's going to impact everything that we do internally, software development lifecycle, our managed services, and, you know, finance.
We are taking the, which also impact, help us in our margin expansion, is, resource management. Is like, you can take it to new heights. I mean, how I can bring the right people with the right skills in the right, offshore, nearshore, onshore, capabilities in the right center, this is something that, obviously taking us to the next level of resource management, which, at the end of the day, you know, we are people, company, and most of our expenses are on people.
And what about the investment from you guys? How much, you know, is there any step up that we're going to have to expect in terms of all these efforts in AI now, or is it-
You know, it's not an AI, and we don't expect any, you know. It's within our current R&D budget. I mean, it's more like priorities with the same budget. It's at the same time, you know, that maybe we are funneling more money to this, we are using this technology to do other things in the R&D in much more efficient way.
Mm-hmm.
So if you look about our spend, it's not just, you know, delivery, operational people, but how we can do. So our R&D machine can produce much more-
Mm.
using this technology, so we are balancing this. And I think that we also have a very good agreement with our partners, like NVIDIA and others, that supporting us in this. For example, NVIDIA, I think that the reason we have such a good partnership because they understand that if they want to double down on the telco industry, which I think only second to the financial services industry, giving them the amount of data.
Mm-hmm.
So, I think that, obviously, we wanted to leverage our technology. They want to leverage our position in the telco market. So it's a good partnership. And again, with Microsoft, we have an amazing partnership, I mean, in different domains.
Just want to make sure if anyone wants to ask any questions, I give them a chance to see someone in the audience who might know something about your company. Okay, let's see.
Shuky, that's a great point about just NVIDIA and the... I kind of wanted to marry that with one of your comments earlier with, you know, only 10% of workflow, workloads in telco kind of being up in the cloud. I think you guys are, what? Around 60% managed services. How does that, you know, the drive to AI, the drive to push more and more of this data up into the cloud, and telco has a ton of it, how does that force their hand in modernizing their systems, their networks, and maybe even pushing more of their workloads into the cloud?
Well, you move to the cloud because of different reasons. By the way, it's not always TCO will not go down because of the cloud. When you move to the cloud, you get to a much more secure environment, time to market is better, and you can do many, many things that you cannot do on the on-premise environment. So it's not just because of and you get, obviously, access to our latest and greatest, which was built obviously on the cloud and not on the legacy platform. So there is many reason to move to the cloud.
I think that if you look about when we work with our customer to move to the cloud, we build for them, like some of them wants to go, like, you know, rip and replace, so we can... Some of them want to do more gradual journey. So, we support them in... But again, what was the question? I'm sure I, I forgot.
Well, just in terms of-
Uh, okay.
-what's gonna...
So the telcos, there is no question everyone will move to the cloud. When I say 10% move to the cloud, I can tell you that half of our large customer are already engaging with us to the journey. So the question should be, what it will be, 2-3 years from now? Probably, it will be a accelerating in moving to the cloud. From our perspective, it's also present also some type of growth opportunity, because in the current environment, you talk about the on-premise environment, we usually will do the operation, and the customer was doing the database operation.... we were, sorry, the data center operation.
It's mainly, in most of our agreements, the customer was owning the data center and was actually managing the data center operation, while we were doing the system operation, which you do. When you move to the cloud, actually, we are taking responsibility also for the cloud ops, which actually replaces the traditional on-premise operation. So it's actually the scope of our managed services on the cloud is broader because we are also responsible for the infrastructure. When it was on premise, we were not doing the infrastructure.
You've laid out a bunch of different growth vectors. If you had to sort of prioritize number one, two, three, in terms of how you want investors to think about what's going to drive growth in the next two years, how would you rank order them, one, two, three, whether it's network modernization, 5G, international expansion, kind of layout for-
It's a, it's a, I get this question a lot. Sometimes it's very difficult to answer because, when we are doing a modernization project for a huge one, for AT&T Consumer, they're doing it to move to the cloud, they're doing it to get 5G capabilities, and they're doing it to uplift their consumer experience. So I, I cannot call out exactly, is it the cloud? Because actually it's the same very significant project that, that answer all of this. I think probably cloud is maybe the number one, and as we said, we said that, it's already 20% of our revenue going double digit. But at the same time, there was a lot of, I would say, project which were 5G related.
By the way, we expect to see some acceleration later next year, because when all our customer will start to have what we call a full 5G or standalone 5G network, today, the radio is 5G and the core network is 4G. So all the new use cases of a quality of service, latency of this, you can hardly get it in very, very small portion of the market. But think, when everything will be deployed, 5G standalone, then you can buy, and I'm sure there are people that will pay the premium to get a guaranteed quality of service, guaranteed quality of latency. So all this. So 5G, I think we're starting to, we're going to see uplift later this year when all of them will be able to deploy 5G standalone.
Most of the AT&T, Verizon, T-Mobile, having a very, I would say, small percentage of their markets, they have 5G standalone. So this is the Holy Grail of 5G. We are not there yet. We see a little bit faster speed, but we don't see yet, so I think that cloud, 5G, fiber rollout, which is all with support and network automation, I'm not sure. That probably cloud is a bit behind on everything, and I think that generative AI will join when we start to see more and more, because generative AI project is not just, you know, the end of the day, what, it's the, all the plumbing, creating the data, preparing it. This is my, by the way, for us, this is most of the money.
I mean, although we give a very holistic, value proposition, because when we do a project with generative AI, we are the prime for, for calling a Microsoft OpenAI, for NVIDIA, for everything. We are the priming, and we are doing it more like a transaction-based type of... But so I, I cannot correlate because in many cases, when, when customer are doing it, there are many reason. I don't think, sometimes we have some project or say, "Oh, okay." For example, the AT&T mainframe deal that we mentioned, this is really only cloud. The idea is to take 45, 50 years old legacy COBOL program move to the cloud. Before all this, you know, mainframe is going away, and there was no way to support mainframe anymore. You can hardly find people who knows COBOL anymore.
But, this is pure cloud, but most of our projects tick the box on some of these,
Thank you very much. I want to thank everyone for coming, and I want to thank Shuky.
Thank you, Shlomo. Good to see you, like always. Thank you.