Unlocking revenue with Network as Code

Ecosystems are the future of the industry but the first step in reaping their rewards comes with the concept of Network as Code. Jitin Bhandari, CTO of Cloud and Network Services at Nokia, explains all.  

Below is a transcript of this podcast. Some parts have been edited for clarity. 

Michael Hainsworth: Communications Service Providers do not want to repeat the past. Over the last 30 years, the telecom industry focused on providing customers with voice communications and later data. It landed the industry the derisive moniker of “dumb pipe”. The smart CSP today has an opportunity to monetize the 5G wireless network in ways we previously never considered. This as the industry builds the next generation networks with virtualized software replacing localized hardware. Nokia Cloud and Network Services CTO Jitin Bhandari tells me the concept of the Network as Code will ensure the CSP of today looks nothing like the CSP of yesterday.

Jitin Bhandari: To understand that phenomenon, we'll have to look back into the recent past and what has happened over the past three decades. Over the past three decades, telecommunications industry has built up layers of complexity and inefficient closed silos network systems while we navigated through the fixed networks, the transition of 2G to 4G mobile technologies, and eventually the convergence of fixed and mobile networks that's well underway. But during all these times so far in every new generation of networks that we know as 2G, 3G, or 4G, our combined inability of our industry to build an open ecosystem has been one of the biggest challenges for us.

And that's what has happened in the recent past if you look at the chapter around 4G, when we were very busy in what I call it in my terms the ‘IP-fication’ of the network, where we were normalizing the network to IP technologies, both for data and voice and fast building and rolling out a 4G network, we became and relegated ourselves to pure connectivity suppliers when a large ecosystem of digital applications thrived on the 4G ecosystem.

Now, the 4G network that we built was very powerful and was built on two foundational principles, that's data and mobility. And we know it all that, as we look at these OTT players, we can name a few from Netflix to Hulus, to Airbnb, and Ubers of the word, they really used and abused the data and mobility aspects of these networks that we built along, while we were busy building 4G and convergence of fixed and mobile, they took the cake from us.

So now the big question for us as we talk about 5G and cloud, two very new interesting technologies, how can we embrace this time around and ensure that we participate in these value chains? That's the crux of it. And I can touch upon a bit more to our listeners because it's very critical to understand. And this topic is very, very near and dear to my heart. If you look at the recent past and I would actually extend and go on and give a very live example which happened in the very near recent history, as I was talking about 4G networks, if you look at the CPaaS industry, the Communication Platform as a Service industry, that actually grew and really thrived from 2013 onwards. Today, if you look at the CPaaS market, the communication platform service providers are simply working on four technology vectors, voice, video, messaging, and email systems.

And what they were fundamentally able to do is built in a horizontal platform that sits on cloud technologies, overrides on this what we call the 4G mobile network. And was simply able to expose these four value vectors to this digital application ecosystems, what we know as of the likes of Uber and Airbnb which are using these basic voice video and messaging communications. And as we know it, that the top CPaaS player is growing 65% year over year. The CAGR of the CPaaS market is going at 35% to a value of 35-billion-dollar market cap. That's quite some phenomenal numbers. And if you think about it, this is just one ecosystem of the digital applications. And we will talk extensively about 5G, but just recent lessons from our history is that in the last seven, eight years, enormous value has been created from on top of the 4G networks. And unfortunately, from a service provider’s perspectives, from the vendor communities' perspective on the telecom industries, we have been somewhat missing to capture this value.

MH: Everyone talks about monetizing 5G, but it's your assertion we've been thinking about from a rather limited way. So how should 5G be built this time? How does seeing the Network as Code help realize the value of 5G?

JB: That's a very, very good question, Michael. And before I go deeper into the construct of Network as Code as we outline it, I want our listeners to understand first why we are talking about Network as Code and where the real value is. We did a recent study and there was a fantastic report that came out of Gartner late last year. And we were partnering with them. And they laid out all the foundations as the new age networks will be built out, all the networks of the future will be built out. And when they stacked up all the B2B spending in this new age networks, it was very clear while there will be a transformation of network infrastructure, adoption of cloud, transformation of the network into more software genome, so on and so forth. And there will be B2B spending all across that. 70% of the relative B2B spending was to be done in the digital application ecosystem.

And this is the digital application ecosystem which sits on top of the network. I'm talking about the cloud infrastructure. I'm talking about network infrastructure that makes and builds the network. And on top of the software in itself, both the network software as well as the operations software. And this digital ecosystem, or 212 billion as they specifically quoted it, or 70% of the value stream of B2B, will be spent in the next five years. So, if 5G is all about verticalization, and if 5G is all about as we recreate these networks, rebuild these networks from the technologies of cloud and cloud native philosophies, how do we tap into these vertical rich ecosystems of digital application is the problem statement in front of us. And that's why we are talking about the value creation from this perspective.

Now, going back to your original question, to connect to this digital application ecosystem, there is also a very rich paradigm that has existed in the last full one decade that has been extensively used by our IT and enterprise ecosystem. And I'm talking and referring to a concept what we call as distributed service chains. If you look at the concept of distributed service chain, it's essentially nothing, but it's chaining two, three, four ecosystem members together to create new value.

Take the example of Uber, we all use Uber. And Uber rides on a cloud network. Uber rides on its own platform. Then Uber connects to the underlying network for using voice and video services. Uber connects to the driver's network. Uber connects it to the rider networks. It's a classic example of a distributed service chain. If you see the digitization behind the pandemic, what has happened into the restaurant industry and how the restaurant industry digitize itself and connected to the digital frameworks of a driver’s network to transform restaurants in throughout the pandemic and not only help them save their businesses, but transform them completely digital, is a great example of a distributed service chain.

So now the question for us is that, as we build this 5G ecosystem, and if we talk about such ecosystems where there is enormous value, close to about 70% as I say it, how do we build up a horizontal platform for communication this time around which enables us to participate in these distributed service chains? How does our network become an integrated part of these distributed service chain is the main paradigm and the crux of this philosophy. And hence, we came up with this concept of what we call as Network as Code.

And Network as Code is all about creating a programmable persona of the network, a programmable persona which is not from an operator's perspective, not from a telecommunications' classic vendor perspective, but it's the programmable persona for all these ecosystem developers and technology application partners who are living and thriving in domains of enterprise, in domains of Web 2.0 and Web 3.0 technologies, in domains of industry 4.0 applications, in scenarios of OT applications, all these third party and first party developers. And how do we create a persona of the network this time around that's much more appealing and relevant to these developer communities? So simply put, Network as Code is our mission to simplify our networks with great extents, provide a persona of programmable nature to these developers' community and technology application partners, and let the 5G network in itself to participate and distribute its service chains. And that's what we call this Network as Code.

MH: So, you're referring to essentially an open API system, the application programming interfaces necessary for developers to build next generation technology on 5G. It sort of feels like Lego. Instead of building a 4G style network with the hardware, we now have components. And we've given software Lego pieces to build whatever we want.

JB: Oh, you're spot on in defining that. And actually, to define these Lego pieces and these components as what you call it, we've got a very specific vision of how the networks of the future will be built around these Lego pieces, as you call it. And what we have realized is, there are five distinctive blocks that are fast emerging now. And these five blocks are what actually is helping us create this persona of what we call as Network as Code. And we can go and talk about it both from bottom-up or top-down. I'll talk about it from a bottom-up perspective, but because typically in the communication industry... And I'm just making it easier for our listeners... we know that there is a fact of life now that these new age networks need to be built on these decentralized, disintegrated cloud infrastructures. So that's there to happen. With the pure promise of 5G of extreme bandwidth and extreme low latencies, you've got to embrace the concepts of edge, far-edge and core data centers.

And one of those building blocks is what we call the scalable decentralized cloud. So, talking about Lego pieces, think of first that as the first Lego piece block. Now, obviously, that layer has tremendous amounts of challenges as we are seeing it because we are talking about three nines cloud infrastructure and building and operating in near real-time networks, which is about five nines or six nines. So, performance, resilience, scale, how do you do unified cloud operations tied up to a digital software centric, critical digital network, that is a challenge itself.

The second block is, what we are going through is, also the transformation of this 5G pieces, what we call the core network, the radios, the transport. And we are doing this what we call the 1.0 era of a huge cloud native transformation moving from physical assets to cloud native assets and how these software building blocks, what I call the digital fabric of the network now, sits over the cloud infrastructure. That's your second layer or the second Lego piece block.

The other two blocks that we believe are going to be very, very interesting are the persona of this network that, as being created, how does it become more business intent and service intent? And this is where the Lego blocks of business intent, service intent operations, highly software-centric in nature, multi-vendor, multi-domain in nature. And that's another Lego piece block that joins the piece of a scalable decentralized cloud to a software-centric critical network. And that's where operations come full to life in this software-centric network. And that becomes another Lego piece block.

And we often then, in the fourth piece, talk about how in this world of full software, where you have software lying at the far edges, edge data centers, core data centers, we are talking about 80, 90 applications across various domains of radio, transport, and core, and also operations, all running in full software paradigm. They are generating huge amount of data, and data and analytics that one could build over it. And that becomes our critical four Lego block.

Now, while I'm talking about these Lego pieces, one thing is critical... And this is just touching up back to the programmable persona of the network... this time around what we have realized is that we have a fundamental change in our building practices and adopting to the cloud native principle that this time around these network or Lego blocks exposes itself in the right way to the programmable persona towards the external developers. And this is where we were talking about the programmability aspect of it.

MH: But CSPs aren't accustomed to opening up their platforms to other players, let alone those who have history of eating their partners lunches. Hasn't the concept of ecosystems and marketplace been tried before and failed? What's going to be different this time around?

JB: That's another fantastic question. And while I touched on the four Lego pieces or the four blocks, the fifth one for us is actually the service marketplace. And I'm having extensive discussions with a lot of CTOs and CTIOs, and it's now becoming a trend where the CIOs and the CTOs are getting unified into one role in our digital service provider trends. And for the last nine months, all those discussions have centered around, how do we enable our service provider friends to build their own service marketplaces? And that's beyond the persona of laying out radios, laying out transports, laying out the core of the network, how do you build operations and analytics? But more from an outside in perspective, if I may call it, because if 5G is all about verticalization, 5G is all about creating new value and connecting to these distributed service chains. We should be front and center talking about how do we create service marketplaces?

Michael, I'll be a bit provocative here, often we have observed and seen, and this is a general market theme now that many CTOs have agreed with me that in the telecommunications' industry, we are a little bit inverted. And what I mean by inverted is, if you look at any other vertical in any other industry, first people built out the business use case, how they will make new money, how the new value would be created? And then the money is invested in building solutions. Whereas, in our industry, for the last two and a half years, we have been spending heavy, heavy investments into rebuilding radios, rebuilding transports, rebuilding cloud infrastructures, and the core of the network, hoping that will create new value over 5G. It's a little inverted way of looking at things.

And finally, it's good to see... It's very encouraging that after this, what I call the 1.0 wave of adoption of 5G and building of the cloud infrastructure, now one of the front and center dialogues is, how do we build up service marketplaces? Now, this service marketplace concept and creation of these open platforms and open ecosystems, we believe this time around is going to be very, very different from what has happened in the past. And let me go and explain to you why we believe this time around it's going to be very, very different. And I will have to use two faces of the coin to explain these phenomena.

One aspect of it is, you have to look at it from purely a network-centric standpoint. You're spot on that 2G 3G, 4G, we are never able to open up the ecosystem. We talked about extensively in 4G. But this time around, if you look at... And I'll go back to a little bit of a technology aspect of things... If you dust off the 3GPP specifications, Release 16, Release 17, and now 5G-Advanced is being talked about, 3GPP specification was built around the persona of building use cases, business intent operations, the concepts of network slicing. So, if you think about it, it has always been an outside in view of the network. If you look at some of the network modeling and the design concepts around core of the network, the radio of the network, it's all driven and defined by cloud native design principles. And one of the foundational principles of cloud native is all about open APIs and programmability of each network function as these network functions are being broken down into microservices.

APIs and abstraction have become mainstream in 3GPP and TMF forums for the last two years. Telco API alliances, GSMA telco automotive alliances. There are many, many forums out there which are talking about APIs and abstraction front and center. Now, that's the network aspect of it.

Now, if you look at it from a cloud, and an IT, and an enterprise perspective of it and we will be seeing behind 5G and cloud the blending of these two technologies. If you look at from an IT standpoint, we are now a decade ahead or what we call the concept of distributed service chains. And we know that the hyperscalers are very rich in marketplaces. So, marketplaces have become a very rich and established concept in IT and enterprise sector. If you look at the cloud infrastructure player now, they're talking about infrastructure as a code concept in a mainstream value. They're talking about programmability of the cloud infrastructure. So, the question that we must ask ourselves from a communications' player, if the cloud infrastructure can be created in the persona of a programmable nature, why not the networks and operations that we are building this time around for 5G?

So, I believe there are many elements to it that are really creating a perfect storm, that I say it. The timing is just perfect, right? This is for the first time we are rebuilding the networks the right way, using the right cloud native design principles, adopting cloud principles, creating that open persona with heavy cataloging and API and abstract ability practice. We believe this is going to be different this time around.

MH: The hyperscalers have a role to play in this. In the past, they were the ones we were talking about eating our lunch. But this time around they are an active participant in building the system. I suppose, one of the big issues that we need to address with them though is going from three nines to five nines.

JB: Yeah, that's an interesting point. And I tell you, we can record a total dedicated discussion about hyperscalers and the role of hyperscalers. It's one of the hottest topics out in the industry. But let's step back, and as I say it often nowadays, it's very clear that there is a trisector that's emerging in the communications marketplace. And when I call the trisector, one clear player of that trisector is the digital service provider themselves. The second is the hyperscalers players who are helping build these decentralized cloud infrastructures.  And the third piece is us, the network vendors, communities like Nokia. And this trisector is an interesting one, and there are a few facets to it.

And you asked me, what's the role of hyperscalers? There are two aspects to it. One aspect of it is about, we believe that hyperscalers have done a phenomenal job in building these scalable, decentralized cloud in the IT and enterprise sector. So, there are practices that can be utilized, used, as we migrate our communications industry on this scalable decentralized cloud. I'm talking about edge ecosystem, far-edge ecosystems. Are there smarter ways of building more autonomous edge sites that can be operated, controlled from central cloud and operations when they are in hundreds and thousands in numbers? So definitely, there are recipes of smartly managing scalability, centralized cloud. And that is where the partnership is going to be interesting.

The other piece we believe where we, as Nokia, are heavily ideating and we have participated in many forums in the industry, and for the last 18 to 24 months from the CTO office of mine, we have had about deep dive relationship with all the major hyperscalers in building out some architectural frameworks of how the new niche networks would build out. One thing that is coming out very clear is the service marketplace. And that's the second area we believe there is going to be some interesting trisectors building up.

Because if you think about it, Michael, we talked extensively about the programmable persona of the network, opening up of the network. And if everything from these abstractions, from all parts of the network... 5G is a very, very powerful ecosystem, as I call it. Because if you look at the Lego pieces of the 5G, whether you're talking about radio, it's giving you proximity and location. Whether you're talking about core of the network, it's giving you voice data messaging. Whether you're talking about operations, it's giving you the ability to slice the network. Whether you're talking about analytics, it's giving you great insights about not only the network infrastructure, and also the operations of the network. And each of these vectors, what I call as the value vectors, can be exposed through a marketplace.

So, the building of a marketplace is also going to be an interesting space or place where the digital service providers will have to actively make front and center strategies. And with a lot of like-minded digital service providers, we as Nokia are partnering heavily... We are having very active dialogues in not only defining and shaping their marketplaces, but also this vision of what we call as a new value creation platform, what we call it for creating new value. So, yes, the debate will continue how this trisector will emerge between the hyperscalers, DSPs, and vendors like us. But one thing, for sure, we believe in strongly is that the control points are with our digital service provider friends. And we will have to strengthen them and their control points so that they build world class 5G networks that are built on cloud native principles.

MH: To your point about the Gartner digital study that shows application ecosystems are going to be valued north of $220 billion, the concept though of Network as Code really hasn't been institutionalized enough within the community. How do we resolve that?

JB: Yeah. And that is going to be an interesting challenge because building communities is not easy as we see it. First of all, you have to realize that as we are building these 5G networks, the software persona and the software cloud native philosophies are going to be embedded in this network everywhere. And I'm talking about, as we are rebuilding operations, rebuilding analytics, rebuilding assurance, rebuilding the core of the network, everything is getting software transformed and so on, so forth. And what we are seeing is, we are seeing two set of developers fast emerging. One set of developers, what we call the first party developers. These are developers who will sit within the ecosystem of digital service providers, and they are operating, managing the network, a software centric network of our CSP friends.

And one of the pieces would be what we are realizing is, how do we build a solid practice of API cataloging, sandboxing capabilities, developer portals? How do we do a very... What do I call it? A developer-centric persona of the network. If you think from a developer standpoint, if you and I were developer say, all we care about if we can get that API that gives me access to a higher bandwidth or an extreme low latency. I don't care about how the network is operating underneath. If I get that API where I can slice that network and get that specific slice I'm using for my enterprise or Web 3.0 use case, I don't really care how the underlying network is operating and working. So, you have to start thinking from that persona of a developer. And that's why more and more cataloging practices, developer portals, abstract ability is going to be front and center as we create these marketplaces.

And the second aspect of it is not only touching the first party developers, but we are also going to open up this ecosystem for third party developers. And these are third party developers and technology application partners who are sitting in specific verticals. I'm talking about mining, ports. I'm talking about industrial automation, all the I4.0 industries and OT applications bring in what we call vertical specificity for their know-how of their vertical. And again, how we appeal to these third-party application ecosystems so that they let us participate in those distributed service chains, the concept that I presented earlier in our talk is going to be critical. So, both aspects, on this digital application ecosystem, the $220 billion B2B spend that we talk about, you want to participate in it, you got to open up the network, you got to provide a developer persona. You got to be more programmable in nature.

MH: And there is time to make these changes possible, not just from a technical perspective, but from a corporate culture perspective as well. To your point, 3GPP has tremendous momentum on the various releases, but we don't have all of those building blocks in place yet because not all of those elements have been released yet.

JB: You are spot on. And, and there is one aspect of it is the technology aspect. But then there is a one bigger aspect that what we are often seeing, it's not easy to build up developer communities. It's not easy to build up vertical technology application partner ecosystems. And there are several efforts now in the communications vertical industry that we are seeing where open labs and open partner ecosystems are being built out. My request and call out... And we are participating in many of those. We are initiating ourselves many such building up of our ecosystems, but my request to digital service provider friends is not only just think of it from a technology lens. Yes, it's important to create that API platform, that monetization platform for these APIs, but how do we focus on changing, a.) our ability to create these developer ecosystems and developer and technology application partner ecosystems? But also, how do we create go market channels fall towards these enterprises so that we can have the right sales and marketing activity from our DSP friend’s standpoint?

That's another piece of it, which is going to be very, very strategic in nature from business value creation, because every single vertical is going to be very specific in nature. And its needs from a 5G network may differ from one vertical to another. So, these vertical specificities and the need from the network is going to be different. Hence, you need to have very specific go-to-market initiatives from that perspective as well.

So, there are both aspects of it, building developer portals, developer communities, and technology application ecosystem partners. That's a critical aspect of it. And the second aspect is, how do I define my go-to-market channels to these verticals? We are seeing that. And the last one, people and skills. That always comes front and center. We are talking about true cloud native networks, adoption of full cloud, a new paradigm of software, new paradigm of APIs, programmable persona. It requires quite a bit of transformation in people and skills as what we know it in the last three decades of how we have built and operated these telco networks.

<< Go to previous episode