Videos
Vikram Meghal on AI-Powered Networks and the Future of Telecommunications
Insights
- Autonomous networks are enabling telecom operators to improve efficiency, resilience, and real-time decision-making across planning, operations, and service delivery.
- AI is shifting from centralized cloud models to edge intelligence, bringing decision-making closer to consumers, devices, and critical infrastructure.
- Telecom operators have an opportunity to become providers of trusted AI-enabled services by combining intelligent infrastructure with data sovereignty and security.
Vikram Meghal, Global Head of Engineering for Networks, Communications, Media and Technology at Infosys, discusses how AI is transforming telecommunications from a connectivity business into an intelligent services platform. He explains why autonomous networks, edge AI, and real-time network intelligence will be critical to supporting billions of connected devices, machines, and applications. Vikram also explores the implications for cybersecurity, workforce transformation, and digital sovereignty, arguing that telecom operators are uniquely positioned to become trusted stewards of data and AI-enabled services. To succeed, telcos must use AI to modernize legacy operations while rapidly building new capabilities that meet rising customer expectations and create new sources of value.
Jeff Kavanaugh:
I'm Jeff Kavanaugh at the Infosys Knowledge Institute here at Infosys Connect Conference in Los Angeles. Very happy to be joined by Vikram Meghal, my colleague who's in charge of engineering services at Infosys.
Vikram Meghal:
Hi Jeff. Good to see you.
Jeff Kavanaugh:
Well, the topic of the day is unlocking value in AI. And obviously it's top of mind for people. What does it mean in engineering services and what's top of mind in AI right now?
Vikram Meghal:
So one thing that most of the network part of the focus is on autonomous networks and that's the way that most of the telcos as well as large enterprises are looking at bringing both efficiency and intelligence into the network. Because as we get ready for the AI world, the intelligence will have to be embedded at different touch points.
Jeff Kavanaugh:
So you're embedding intelligence at various points in the telecommunications network?
Vikram Meghal:
In the network itself. And that will actually give rise to a lot of edge AI use cases.
Jeff Kavanaugh:
This frontier model kind of thing?
Vikram Meghal:
Exactly, exactly. So right now, AI is more focused on cloud-delivered AI, which is have a big hub, and you deliver inferencing as well as...
Jeff Kavanaugh:
it’s an improvement over the old days, but it's only part of the way. Part of... Yeah.
Vikram Meghal:
So once you look at really getting into edge AI and touching every point of life that we live, then it gets into more how do you embed that intelligence closer to the consumer and that's where the network comes into the picture. Because what's between the cloud and the consumer is the network and the network needs to get more intelligent as we go forward.
Vikram Meghal:
So they are actually looking at how they can build a layered approach to bring intelligence into their products, but also ensure that this intelligence can be shared with other products from their competitors or partners. That is very important. They are happy to be more open. They have to be more open. And TMF is driving some of that structure.
Jeff Kavanaugh:
What is?
Vikram Meghal:
TMF, the TeleManagement Forum. And they are ensuring that finally products can talk to each other at an intelligence level rather than just at a protocol level.
Jeff Kavanaugh:
So standards have to come along far enough so that you can't have these competitive moats. They have to talk to each other.
Vikram Meghal:
Yeah, and that's where the telcos are playing a very important role because while we have this embedded intelligence into different products, telcos don't sell products. They actually create services on top and that's what earns their revenue.
Jeff Kavanaugh:
And they have enough scale they can be a bit of a channel master and kind of influence them. Influencing that ecosystem.
Vikram Meghal:
So that's exactly the role. So we are actually working with telcos in ensuring that they can step up their autonomous network preparedness, go from level two where most of the people are to level four. So that journey is going to be probably a two to three year journey. And we are actually actively taking people from autonomous AN level two to AN level three plus. And that's where most of the investments are going in.
Jeff Kavanaugh:
The average consumer, they understand what they're able to do on a smartphone or on a machine. The enterprise departments, they can kind of know what their capabilities are. What does it really mean to go to that next level for a network autonomous?
Vikram Meghal:
Yeah, so there are two ways to look at it. One is internally, how do they manage these networks? And that's where network planning, network engineering, network operations will completely get disrupted because it will get more real-time, more autonomous so that you can react to changes in the network very quickly.
Jeff Kavanaugh:
It's hope it’s helping bad things not happen.
Vikram Meghal:
Not happen. What about the good things happen? Yes. And it will also allow people to, from a consumer perspective, let's say you are having a very critical conference call while your kids are watching a movie.
Jeff Kavanaugh:
Right. Yep.
Vikram Meghal:
And you want to ensure that this traffic gets prioritized over the gaming or the movie watching that's happening in your environment.
Jeff Kavanaugh:
So it isn’t just a big pipe, you can actually prioritize?
Vikram Meghal:
You can actually create that intelligence so that you can sense that there is an important activity happening. The network actually reconfigures itself to get ready for this new stream which is coming in which is high priority. But to do that effectively, you need to build intelligence into the network. You need to build intelligence into the edge where the customer premise equipment sits and then throughout the network. And that is the important part, which will hit the consumers.
Now we are just talking about consumer-initiated traffic. But tomorrow, there will be lot of automation-initiated traffic.
Jeff Kavanaugh:
I’m thinking about the industrial internet.
Vikram Meghal:
Exactly. Machines.
Jeff Kavanaugh:
Machines, vehicles… Yeah exactly.
Vikram Meghal:
Vehicles, machine, and every instrument that you can imagine.
Jeff Kavanaugh:
Billions of connected devices. Exactly.
Vikram Meghal:
So that's when the ability to sense this intelligently and react to that at real-time is going to become very important. One of the bad things that, this is a good part of it, the bad part is what is happening with Mythos and how...
Jeff Kavanaugh:
We’re about to get to cyber in a moment, let's talk about security.
Vikram Meghal:
Yeah, so actually if you look at it, networks are becoming a critical infrastructure for any country. And that's where the rate at which these are going to be attacked in the future is going to go up. And you need to have that ability to automatically isolate certain sections of the network to understand what is causing this attack or impact, right? And that is what, again, intelligent networks will be able to do. Rather than having a human-led interaction or intervention to isolate some of those components. How do you automatically isolate certain sections of the network while you analyze what is causing that attack and then quickly react to that? So that complete change means you need more autonomous functions into your network. And that's exactly what building intelligence is all about.
Jeff Kavanaugh:
Talking about the tech side, the consumer side. What about the employees? As all these agents and all this capability comes in, is it a threat to their employment? Do they get to have more capabilities? What's going to happen?
Vikram Meghal:
The one thing is that their ways of working are going to change. And that's where what we are looking at is how do you take the tech plus ops together, rather than just building beautiful tech, which never gets adopted, right? So the focus is on network operations, for example. How do you do both the technology and drive the adoption at the same pace? And that means, it's not only about adopting tools, which used to be what automation did. This is about reimagining the way you do the work. The question to be asked is, does a human need to do this? Or can it be delegated to a machine or an agent? And that is driving the change.
Vikram Meghal:
So that reimagination is going to actually bring new roles into the, for example, network ops area. It's obviously going to challenge some of the existing skills. People will have to build new skills there. And over time, it will actually help the humans to focus on more of the creative part of it on how do you add more security layers, for example, rather than fixing problems which are mundane, right? So that's where I think it's going. It will impact the skills part, people who can actually write this, learn from this, and adopt this faster will actually have a great future ahead.
Jeff Kavanaugh:
Let’s talk briefly, because you and I have known each other for decades, for a couple decades and the operating model for telcos and everyone in between has shifted from being very profitable to fighting for existence and you know where is the value created, where is it captured, who gets to keep it? Who are the winners and losers going to be in the next few years?
Vikram Meghal:
I think the telcos which are going to look at AI as an opportunity to provide more services, for example, lot of countries are looking at sovereignty as a very important aspect, right? And telcos are the natural nation builders in these smaller countries.
Jeff Kavanaugh:
They’re a utility now of some kind.
Vikram Meghal:
Exactly, and they are a critical utility. Exactly, and they are a critical utility. Yes, you’re right. An intelligent utility. Yes. So I think that is a natural path that telcos can take because if you look at the entire IoT era the over-the-top players took all the revenue and the money and the telcos saw very little of that and I think this is the time when telcos can actually focus on how do they build meaningful intelligence which can be priced above the normal.
Jeff Kavanaugh:
And they’ll have political support because the sovereignty comes in.
Vikram Meghal:
Exactly, and that's where I think everyone is getting smarter about how their data can be used and how it can be restricted. And the telcos have proven to be the keepers of critical data over the ages. So I think that’s the role they need to enhance and focus on, and that will lead to more revenue accumulation for them.
Jeff Kavanaugh:
Given your role, and your vantage point with all these things that are going on at Infosys, what's the one thing that telco leaders and network leaders should think about in the next nine months or so to get through this?
Vikram Meghal:
I think they have to look at a three-pronged strategy. Number one, they have to embrace AI to ensure that they can take out costs from network operations, planning design, running their existing systems. Because they have a lot of legacy because network technology was built across generations. So you just went on building one above the other, right? So that legacy is not going to go away, so they have to use AI to ensure that they can manage this very effectively, right? But at the same time, they have to think about building new capabilities at the speed of customer expectations. See consumers are expecting things to happen immediately because they are used to that in their day-to-day operations, or day-to-day life. Now if telcos can't match that pace, then they will actually look for other options and which will actually mean erosion of telco bandwidth. For example, OTT TV actually taking over what used to be paid TV, right? Just because the pace of change was so slow. So I think telcos have to have this dual speed strategy. One is take the cost out of the legacy, manage that in a very efficient way, but build the new very rapidly with the agility that AI makes it happen.
Jeff Kavanaugh:
I hear duality, a phrase coming up a lot, bimodal duality. Listen Vikram, you gave us a lot to chew on here. Thanks for your insights.
Vikram Meghal:
Great talking with you. Thanks, Jeff.
Jeff Kavanaugh:
Thank you. Pleasure. I'm Jeff Kavanaugh. Until next time, keep learning and keep sharing.
