Clarence Reynolds, TelecomTV (00:09):
What role will network APIs play in creating new revenue streams for telcos? Chivas Nambiar, General Manager of Telco Business at AWS is here to share his insights on AWS's role in accelerating network innovation, 5G deployments and generative AI in telecom. Chivas, thank you for being with us today,

Chivas Nambiar, AWS (00:30):
Clarence. Thank you for having me on. It's exciting to be part of this conversation and I'm looking forward to this discussion.

Clarence Reynolds, TelecomTV (00:38):
AWS recently announced a major cloud migration deal with Telefonica Germany. Can you share more details on what this Brownfield operator migration entails and why it's significant for the telecom industry?

Chivas Nambiar, AWS (00:52):
Yeah, we're very excited about this journey. If you think about Telefonica Germany, they have had for a little while now a vision of what the network of the future needs to look like, and we've been partnering with them and their key ISV, in this case Nokia, to envision what the future looks like and to get that deployed. So in this particular case, we've got a deployment with Nokia's Cloud Native Core 5G core architecture, which is a completely software based packet core, right? It's running on top of AWS's cloud in our AWS regions and our availability zones, which are think both them as data centers that run in our regions. And we've built it in such a way that it's actually supports the stringent requirements of a commercial telephony operator. Now, in terms of what we ended up building with Telefonica and Nokia, they're using critical AWS capabilities including things like AWS Elastic Kubernetes service, which is a container orchestration service that we provide, which is a fully managed service.

(01:53):
It takes all of the container functions, the containerized functions that Nokia has. You run it on EKS, and then you use Amazon's Elastic Cloud compute, which our compute infrastructure to support different workloads associated with it. All of this running in a way that Telefonica has figured out how to run it with the right kind of resiliency to operate a commercial network. And it's the first time we've seen a brownfield operator work with an ISV and AWS to deploy something like this at scale. One of the key things that we think long-term this also allows Telefonica to do is they look at the infrastructure that is built and some of the key capabilities around security in this public cloud, which uses AWS's lightweight, secure hypervisor called Nitro, and they've figured out how to use it to protect this workload and now can figure out how to take that and run it in regions, run it closer to the edge where they need to. So we're excited about what the long-term model of this network of the future looks like with Telefonica.

Clarence Reynolds, TelecomTV (02:54):
As you know, 5G networks and services are currently a key focus for a telecom providers. How is AWS enabling telcos to accelerate their 5G innovation and deployment through its cloud services and partner ecosystem?

Chivas Nambiar, AWS (03:09):
Clarence, that telcos run some of the world's largest, most complex networks. They're critical national infrastructure in most cases. So as telcos have kind of worked through their 5G deployment, it has been rewarding for us at AWS in particular to work with some of the most innovative telcos I mentioned Telefonica. If you look at Japan and N tt doomo, which is one of the largest carriers there, dish in the United States, which is a brand new network that's spun up. All these folks have really spent time with us to look at concepts that were developed in the cloud and then figure out how to take those concepts and deploy it in the telco ecosystem. And that helps them really start to envision what the telco of the future looks like. This includes everything from speeding up deployments, simplifying operations, taking all the data that is generated in these networks, and then using AI and recently gen AI capabilities to optimize the running of these networks.

(04:00):
As folks who have lived in the industry for a while know most of the complexity and the cost really lies in the long-term. Running of these networks and automation and data-driven but safe kind of mediation are two areas that the telcos we work with are mostly excited about. The second part of this is that a lot of the innovation in telcos really comes from the ISV ecosystem, right? So the ISVs and the NEPs and telcos are very interested in figuring out how to drive ever short cycles between a feature being envisioned with these ISVs and partners that they have, and then being able to take that and bring it in front of their customers as fast as they can, but with the right kind of controls and guardrails that automation provides. And if you look at public cloud providers and AWS in particular, we with our largest partner ecosystem of more than 130,000 partners really have figured out how to do this at scale. And that's what we're working with telcos on.

Clarence Reynolds, TelecomTV (04:56):
Generative AI is becoming tremendously important across industries. What specific generative AI services and solutions is AWS providing to help telecom companies unlock new capabilities and business models

Chivas Nambiar, AWS (05:09):
Models? I haven't had a customer conversation where we haven't touched on gene AI in I think about a year and a half. So it is top of mind for everybody that we talk to with Gene ai, though we're finding that our telcos really want to think about in different buckets. So we kind of think about it in three buckets. The first one is for telcos that have the skillset and the capability and the desire to go build models or really control all the knobs of the gen AI ecosystem that they want to build. AWS provides a deep set of capabilities around things like SageMaker, which is our way for you to take models or build models to train models and then tune them with your own data in your own environments in a safe and protected fashion. Underneath it, we provide access to GPU and other accelerated infrastructure, including ability for you to use AWS custom silicon to drive down the cost of your training through training and chip sets and cost of your inference through in friendship sets that we're continuing to work on.

(06:10):
So overall, that underlying ability to take and build models and really touch the hardware infrastructure, we think there's a bucket of telcos that really want to spend time there. For the majority of telcos that I work with though, what they really want to see is in a world where all of these models and the evolution of the models is really fast and you cannot make decisions today that last for two to five years out, which has been the traditional model. When we pick ISVs vendors and partners in the past, they want a fully managed environment where they can build one set of guardrails, one set of protections for their data, and then be able to bring in models that are unique for each use case, fine tune them with their data in a safe fashion. And then over time, if they find that other models are better or they find smaller models that are better or find better models, they can go swap 'em out without actually having to rebuild everything.

(07:03):
And so the Amazon's bedrock service is what they've been using for it. It's built in a way that is a fully managed experience. You can use and choose a number of different models for your use cases. You can fine tune with your data, you can put the right set of guardrails around it. So that's really resonating with our customers who want to use genai. Then finally, for some particular use cases, there are GENAI enabled applications, and recently we talked about British Telecom enabling their developer audience to go out and use what is now Q for developers, which is a ability for them to infuse gen AI into the ID as they're building their products. And that has proven to be very useful for British Telecom. They have found success to the point that they've decided to roll it out across a large group of their engineering teams and in the first few months of usage, and it generated about a hundred thousand lines of code all by itself, and this takes away a lot of the undifferentiated heavy lifting. So we think there are a few of those use cases where they will look to AWS to provide that kind of applications that are on top. And in some cases those applications will come from us and some cases they'll come from partners. And partners are obviously using all of the same capabilities that I mentioned, that our customers are using lots happening in this space.

Clarence Reynolds, TelecomTV (08:22):
I see that AWS is also participating in the network, API panel discussion, and there's a growing momentum around exposing network APIs to unlock new revenue streams. Can you share more about the role AWS and other hyperscalers are playing in enabling this and how can AWS's services help telcos deliver on the monetization promise of this API driven approach?

Chivas Nambiar, AWS (08:48):
Yeah, for decades, we've been talking about how do we take some of the capabilities we have in these large complex networks we run and then be able to provide that as an API to end users, ISVs and customers. And we've struggled a little bit, and I think the struggle has been that if you think about how these APIs are created and deployed in these networks, then you have in every country, let's say two to four operators to get broad enough coverage across the country. All of them have their own implementations based on whatever ISVs and partners they're using. And then how do you aggregate all of that together to provide one single seamless developer experience? That's been really hard to do, but over the last few years now, especially with some of the industry work around things like kamara, we are starting to see standardization of these APIs.

(09:39):
And now you have an opportunity to take these standardized APIs, expose 'em, provide access to all the network functions we're talking about, and then start to provide that to a large level of audience. And so if you look at AWS, we have the largest developer and ISV ecosystem that has worked with APIs. And so developers really want that frictionless and a compelling experience that allows them to access not just the telco APIs, but in parallel also all of the cloud capabilities, the storage, the compute, the AI capabilities to then take advantage of the network APIs. And I think that's what we're most excited about. So we've been working with a number of our customers, and we wrote a blog about this actually that we surfaced at NWC this year where we talked to folks like Orang and T-Mobile and Liberty Global and said, Hey, what could we do in terms of building this experience for our developers? So Derek, they're partnering with us and we're excited about how this is coming together. We will likely have more to talk about over the course of the year as we build it.

Clarence Reynolds, TelecomTV (10:44):
Chivas, this is your first time attending the DSP Leaders World Forum. What are your expectations coming into this event, and is there any particular topic, area or potential collaboration that you're most excited to explore with the attendees?

Chivas Nambiar, AWS (10:59):
I'm excited to be part of the forum, just the opportunity for us all to sit down together, learn as a community with some hard earned lessons that we will all share, and also having a venue where we can mull over the future of the telecom industry gets excited about it. I think that's super important. These are the type of venues that build communities of leaders and practitioners that really help shift and change the industry and define where we need to go. We've long elaborated on things like standards, but I think what's different now in this wave of transformation is it's no longer about what we're going to do, which is what the standards used to define, but how you do it and how fast you do it. We were just talking about APIs and when we talk about things like how do you translate from APIs being something you standardize and you have it as an offering to how does it actually effectively become useful for the telcos, which is monetization?

(11:49):
Then you start to have conversations with folks like AWS who built large marketplaces where we have the models and the mechanisms that allow you to monetize these APIs, right? So I think that that conversation with a good group of people that have operated across telcos, across hyperscalers, across some of the key ISV partners will get us to a place where we come out of this with some exciting new things to go figure out how to build. We come out of this hopefully with a sense of possibility about what we can build, and in the end, I think it's a question of how do we go out and build away for us to take all the undifferentiated heavy lifting out of building these complex ecosystems and help translate it into deeply personal experiences for our customers. Personally, for me, I think as I meet with the folks in the DSP forum, I am excited about connecting with some change agents that we can work with because a lot of this is fun about working in the industry and fun about working with people who are forward looking and willing to take some risks and go build some things together.

(12:55):
So that's probably for me personally, the biggest thing I'm excited about.

Clarence Reynolds, TelecomTV (13:00):
Chivas, it's great to talk to you today and thank you for your insights,

Chivas Nambiar, AWS (13:04):
Clarence. Thank you for having me on. I look forward to seeing you there.