Transforming Network Management with Forward Networks’ Nikhil Handigol

This is Textron tv. Hey guys, thanks for the throw. We're here, Whitney Kill Al, who is, uh, co-founder of Forward Networks.

And we're talking about how digital twin technology will be applied to the way we manage networking. And it's a lot more than just putting together a bunch of pretty diagrams. Hey, Nikhil, welcome to the show.

Thank you, Michael. Thank you for having me. Walk us through exactly what's happening here.

'cause I think we're looking at a fundamental change in the way we think about managing networks, but um, that's not clear to me that people have put digital twins and networking together just yet in their minds. So what's happening here? Yeah, absolutely.

Uh, so let me start with the goal. So our goal is to make networks, the enterprise networks of the world reliable and secure. And the way we are doing this is through a technology called Digital Twin.

And uh, the best way to think of this is through an analogy. Think of it as Google Maps for your network, Google Maps or, or any other, uh, mobile application for maps. GPS application, uh, on your mobile phone is, is basically a digital twin of the world's roads.

And we are building a similar technology, but for the enterprise network, How does it learn what's going on in the network? 'cause a lot of people would point to Google Maps and they get frustrated 'cause, well, the last 50 yards or so is never quite what it's supposed to be anyway. So how do we kind of get to this more accurate representation in the network that seems to be constantly changing?

Yeah, that's a great question. So the way it works is it's a software platform that basically connects to all the devices in the network. Your, uh, on-prem devices like switches, routers, load balancers, firewalls, your cloud resources like your VPCs or V Nets and collects configuration and dynamic state from these devices.

So layer two, two, layer four, uh, and even beyond and builds a behaviorally accurate model of the network. That's the fundamental technology. It's a behaviorally accurate model of how this network is behaving.

Do I have to install some sort of agent software to make that happen or can it just collect all that data that it needs on its own? Yeah, that's, that's the best part, right? So you, uh, this works on your networks as they stand today.

You don't require any agents to be installed on those devices specifically. Uh, the network itself doesn't have to, uh, comply to a particular architecture. It just connects to the devices just like a regular operator would, right?

SSH or CLI. Uh, so, but its software that's doing that and collects information that it needs. Can I make a change to the digital twin that would be reflected in a change in the physical network or is it more right now just capturing the physical network?

But I wonder, you know, are we on a path towards using the twin to actually manage the network? Yeah, That's, that's a fantastic question. And the short answer is today the digital twin technology that we have built is primarily read only.

So it collects information and provides the insights to the users. But we are making, we are, uh, we are quickly making progress towards a day, uh, not far in the future where we will be able to make changes in the digital twin and see the effects of those changes within the digital twin before actually applying them on the network. Of course, we hear a lot about all things AI these days.

What is the intersection gonna be between digital twins, ai and the way we think about managing networks? Uh, uh, again, another fantastic question. Uh, we believe there is a strong synergy and, and a very complimentary, uh, nature here between the digital twin technology and ai.

In fact, we believe that digital twin technology is going to be the enabler of AI for networking and security. And there are two reasons for this. One is data.

Like, so AI systems are only as good as the data that they are, they have access to that they're built on. And digital twin is, is, uh, probably is one of the most, uh, comprehensive repositories of some of the most unique data about the network. The second aspect is trust.

So, uh, when you use ai, uh, I think most people use chat GBT these days. You get an answer. You don't know, uh, uh, how do you trust the answer that is coming from there?

So apply the same thing to networking. If the AI system is making recommendations for particular changes in the network, how do you trust that answer enough to be able to implement those changes, implement those suggestions on the production network. And again, this is where digital twin comes in, uh, because validation or verification of the network's behavior is a fundamental characteristic of the digital twin.

So, uh, data and an ability to trust the ai, these are the two reasons why we believe digital twin is going to be the key enabler, uh, of AI for networking and security. Do you think networking will become more responsive? And I ask the question because we see, uh, somebody will provision an application and the virtual machine that they spin up is measured in seconds, if not minutes.

The network, on the other hand, changes are typically can take anywhere from hours to days. Um, so are we getting the point where I can manage the networking environment as jointly as I manage the rest of the infrastructure environment? Fantastic question.

I think this is, this is a problem that is ailing every single industry, uh, out there. It's the networking is the bottleneck today, as you said, you can spin up applications really, really quickly, but how do you provide the connectivity for the application while ensuring that the entire security posture is tight? This is a problem.

And with the digital twin technology, with an ability to understand how the network is behaving before you make the change and after you make the change, I think this is going to be that catalyst to make networks agile. So this is a part of our mission to make networks agile while keeping them secure and reliable. Uh, so I think digital twin technology today and as it evolves, especially in conjunction with AI capabilities, is gonna be a key enabler for making networking as agile, as compute and storage.

How programmable is all this gonna get? And I asked the question because sometimes folks want to manage infrastructure's code, other times they want to use some sort of GUI with a, um, a software defined networking platform. And then maybe it's the network administrator who does that.

But, um, it seems like digital twins are taking that whole conversation up to another level. So ultimately is the network becoming a piece of software rather than hardware? I think that's the way this is going.

Uh, in fact, uh, I think, I think automation has to be a part of this story. There are no two ways about it. And the way you enable automation is by provide, by building APIs, by providing APIs with the, the tools and platforms that you create.

And same goes with digital twin as well. I mean, especially the platform that we have created here at Forward Networks. Uh, it comes with full API and several of, several of our customers actually use automation systems that interact with the digital twin through its API I'm gonna assume that that all means that maybe I don't have to be as much of a specialist to manage the network as we go forward.

Will the networking function get more integrated with the rest of the IT function? I mean, will I need as many networking specialists as I have today or can I allow mere mortals to manage the network? In fact, we are seeing some of this already with our customers.

So what, what the digital twin does is it breaks down this knowledge silos because we have the digital twin fundamentally takes the data from these network devices and converts it into a vendor independent model, a vendor, independent technology, independent model of the network. So what it does is it's, it raises the level of abstraction, which means that you don't need, uh, uh, a person who is probably not a super expert in one particular vendor. Technology can still make sense of what's going on in those devices.

So you don't have these knowledge silos anymore. I think this is where, this is how it'll eventually evolve where we basically raise the level of abstraction that's needed. For example, if you're a software developer, you don't need to understand how the assembly code works.

You just operate at a high level of abstraction. I think the same will happen with networking eventually. Has networking reached a point where it's just too complicated for anybody to manage without the help of some sort of machine, whether it's AI or software or whatever the tool may be?

I'm asking the question because, you know, there's thousands or maybe hundreds of thousands of endpoints and they're all talking to some sort of, uh, server at the edge gateway and that in turn is talking to data centers and assets in the cloud, and all of that is connected by a network. But I don't know if anybody really understands where all those connections are being made. Yeah, I, I, I a hundred percent agree with you.

I think networks have reached a stage where the, the amount, the sheer amount of complexity has gone beyond human capability alone. I think human capability fundamentally needs to be augmented with smart software and that's what we are building here at Forward Networks. Do you think also that we don't appreciate the laws of physics when it comes to networking in those environments?

'cause all these applications we have seem to be increasingly latency sensitive and, um, there's an old joke that says, you know, what's the one thing that an IT admin and a developer can agree on? It's the, it's the networking guy's fault. Um, how do we kind of like get to the point where I can see enough about what's happening on the network to make sure that my applications are meeting, uh, SLA requirements, quality of service?

It seems like it, it's become more challenging, uh, With increasing complexity. I agree. I think there is, there is an increased risk, but I don't think there is anything fundamental preventing from networks being, uh, being tamed.

I mean like the, the complexity of the networks being tamed to a level where we can have confidence that the network is doing as what it should. And it is secure in terms of its, uh, in, in terms of it, its attack surface in terms of its security posture. Uh, it's entirely possible, but we need, we do need smart software to augment human capability.

And how do we get there from where we are? Do we need to rip out all the routers and switches we have today? Or is there some other way to think about this?

'cause it seems like, um, it might be a heavy lift, but I'm not sure. Yeah, I think, uh, I mean, rep ripping and replacing is not an option really, right? Uh, so the, the only, uh, I think the only practical path here is one that is evolutionary in nature.

So this is where a technology like the digital twin technology comes in, which operates on the networks as they stand today with zero changes, right? So the, the architecture that you've got, the, the multi-vendor, uh, system that you've got, uh, right with with the devices that you've got. So basically use smart software to augment human capability and gradually evolve it to whatever, uh, whatever the destination may be.

But the path has to be evolutionary and this is why, uh, it's important for us to, uh, understand this reality. So what's your best advice to folks about how to get from where they are today to that emerging new reality? 'cause I, I think a lot of times I can't figure out if it's a technical issue or a cultural issue where people are just gotten into a habit of managing things the way they always have.

Uh, there's definitely a process issue. Uh, I think, uh, people, uh, especially at leadership positions need to think, think about what the processes need to be so that, uh, there is efficient communication, there is collaboration and all those, uh, layer eight, uh, you know, uh, as they call it, layer, layer rate, uh, issues that need to be addressed at the same time. I think there is, uh, there is a lot of, uh, low hanging fruit to be plugged at, uh, at the infrastructure, at the technology level.

And this is, uh, digital twin is a great example. Uh, it's a great example of technology that you can implement today and gain benefits from day one with zero changes, gain the confidence that uh, your network is doing what it should. Alright, folks, you heard it here.

There's a lot of data on the network and there's a lot of APIs. The question is, is how to harness it using something that doesn't require me to be a programming genius to work all those low level APIs. Hey, HIL, thanks for being on the show.

Thank you, Mike. And back to you guys in the studio.