Why Intelligent Connectivity Is the Next Frontier for Physical AI

Hey everyone. Welcome back here to Techstrong TV. My next guest is the first time here.

I'm excited to have him on. His name is Paymon Samadi. I hope I got that right.

Paymon, welcome to Techstrong TV. Thanks, Alan. Thanks for having me.

Did we get your name right? If not, please correct us. Yeah?

It is correct. Paymon, yes. Okay.

Paymon is the CEO and co-founder of a company called Aeno. Yes? I like to get my names right when I can.

And he's going to tell us all about Aeno, and they've got a brand-new product in AI agentic network observability. We're going to get to that. But I wanted to start off talking about Paymon.

As a CEO co-founder myself, I love meeting other founders and hearing their story. Right? Because founding a company is not something you do lightly or half-heartedly.

Right? It's your passion that carries the day. Talk to us a little bit about your journey, Paymon, and what led you to co-found the company.

Certainly. So my background is both on the academic side and also industry. I studied photonics, got a PhD at McGill University, and there I joined a company as a product manager, doing something kind of similar to what I do now, how to simulate optical networks as a whole.

And then I missed doing research, so I found a research position at Columbia University. I went there for four years, and I was working on how the future of networks should look like with all the boom that's coming with data. And did a lot of collaborative work with different companies from Nvidia to Microsoft to AT&T, and then it was at the starting point that AI ML was getting more practical, and we thought, yeah, this can enter telecom world and connectivity world because this is the future.

Absolutely. And I've learned over my business career, sometimes it's better to be lucky than smart, and it's good to be both. But talk about being in the kind of right place at the right time.

This whole AI thing has just blown up. So what was the passion behind Aeno? What led you from doing what you were doing, you seemed like enjoying the research, being on cutting edge, to throwing your hat in the ring?

Yeah, we saw some changes. We realized everything that is connecting to networks is evolving rapidly. From the way we use our phones, all the way to the industrial settings that we have remote control devices, autonomous system robotics, and now physical AI.

And that means networks are becoming as critical as they could ever be, but it also means we are at this new phase of building a lot of new networks and upgrading the existing ones in the next decade. And then when we look at the workflows that we have and how we run, build, manage our networks all together, we realize that has barely changed in the past 10, 15 years, which we saw a gap. We realized that this gap's going to really hurt us because now we need to have networks everywhere very fast.

And we thought this is the best thing that we could do. The other part that is actually very interesting is that one out of four people who work in connectivity world's going to retire in the next five years. And if you go to any connectivity conference, unfortunately, there's no inflow.

Everyone's over 40, 50 years old. And that's where the role of AI, physical AI, and agentic AI can really come in, and that's how this company started. Really?

What a great story. I used to think that about the mainframe market, right? You go to a mainframe, open source mainframe, or just a mainframe conference, IBM, whatever, and it seemed everybody there at one time or another worked at IBM.

Right? IBM was the fountainhead, right, of the whole... And it is.

It's the main hardware. But everyone was my age, right? But a funny thing has happened.

I don't know if it was from COVID, or maybe it started before COVID. But people need COBOL because all those mainframes aren't going away. They're still at the heart of so many big enterprise businesses who have billions of dollars invested in their mainframes.

And we need these programs, and we need to make modifications to these programs. And even though today's mainframes are so modern, they run Java, they run Rust, they run so much, but they run a lot of COBOL programs. And now when you go to these, you're starting to see younger people because they realize, A, there's jobs there, there's work to be done.

Right? There are more COBOL programmers in India, for instance, because it's a guaranteed job for them, right, in a very competitive market. So look, maybe the same thing happens here in connectivity.

This is going to be an interesting thing. Or does AI just run the table and make it, it's okay, we got it. So how long have you been-- When did you found Aeno?

The company was founded in 2018. Our first product was on spectrum optimizations, for telco operators. It was at the start of the 5G.

There was this hard answer of where to deploy spectrum, when to deploy, and then how much to deploy. We started that, and then in 2023, we realized this new gap, and we thought this would be a good pivot for our companyWe launched our new product in end of 2024, and then 2025, we gain about 40 customers from all major SIs, equipment vendors, some enterprises. And the whole idea about our company is this concept of intelligent connectivity.

Mm-hmm. We realize for the next decades, for all these robotics war and physical AI to work, we'll need new workflows to manage the life cycle connectivity there. And it's based on three pillars.

One is that we want a unified platform across the life cycle. You cannot have one tool for your design, one tool for your validation, one tool for your monitoring, one tool for your asset management, one tool for your IoT. Also, you don't want to have one tool for your Wi-Fi, one tool for your DAS.

That means you need different tools and different teams, and that's not scalable. So let's unify them in one platform for all types of networks that you have in the entire life cycle. Second, anywhere that you build a network, you build it for some physical space.

So it's very important that have a very accurate representation of that physical space. So we find very creative ways to create 3D digital twin of the environment based on if it's indoor, if it's outdoor, we have different methods to do that. And based on where you are in the life cycle, we build a digital twin that serves that part with the cost and the precision that is required in that phase.

If you do feasibility study, it's very different than if you're monitoring a very complex mission-critical network there. And the last part is that we need to use agents. We have sales engineers, design engineers, NOC engineers, network management engineers.

All of those could become an agent. And you're going to have this 3D digital twin, you're going to have the army of agents around it that are interacting with it. But on top, the human becomes the general.

It's sitting there as this army of agents and managing this whole system in front. I was over at the RSAC conference in San Francisco the last couple weeks. I don't know if it was last month or this month.

I'm confused already, but it's a couple weeks ago. And a term that we were hearing there a lot, because it was all about agents and agentic, the human in the loop. Mm-hmm.

But here's my take on it, Peyman. That the velocity and scale of these AI agentic agents that are doing these jobs is so rapid, I don't know if we could have a human in the loop of everything. I think we've got to take the humans up a, let's call it human at the helm.

Yeah. Right? So it kind of oversees it all, but it's just not in every workflow step here, right?

And that's the only way it's going to scale, right? Yeah. And I think that's exactly what you're talking about here as well, is more, as I say, human at the helm rather than human in the loop.

It's an interesting thing. When I first launched Techstrong 12 years ago, APM, application performance management was all the rage. I have a good friend of mine, Dave Link, who started a company 20 years ago, more than 20 years ago, 23 maybe, Science Logic.

It was network management, basically, right? Monitoring and management. All of these things are now called observability.

Observability is a big word. It covers a lot of different things. I understand that you guys have recently announced an observability solution.

But that could be, who knows, right? Let me hear from you. What's this new solution, and how does it help people?

Certainly. So, as I mentioned, we had the 3D digital twin. We had a design platform that says how the network is supposed to perform.

And then after that, once you get to day two, you need to observe the network. And there are tools in the market. But we saw a few gaps.

First thing we saw is that a majority of the tools are focusing on a specific type of network. It does Wi-Fi, it does DAS, it does private cellular, but the networks are converging. Every enterprise is going to have multiple different types of networks.

So let's have a platform that you could monitor and observe all types of networks that you have. Second thing that we realized was that majority of the tools are either looking at the network from the hardware point of view, like how the network sees things are happening, not from the user experience. And there are some tools that are looking from the user experience side.

But there are many different pieces there. What we thought was that you put some probes, you see everything from the user experience, but in order to automate and allow AI to do root cause analysis, let's give it the 3D digital twin, tell them what's the physical characteristics of this location that the problem happens. Second, tell it from our design, what is the predictive performance, and then all the monitoring and observability that comes from our sensors.

And as a whole, allow AI to do a much better job because it has way more context there than some telemetry data that comes in that it has to figure out, which is a very hard thing to do. So that's the big news. We have basically launched this new observability product.

It is part of our intelligent connectivity platformSo for every location, you have a 3D model, you have a design, and now you have the monitoring side. It's fully AI native, in a sense that the AI defines what is a good performance for this network, what are the things that needs to trigger an incident. And then once those incident happens, AI makes a decision, what are the type of playbooks and tests I need to run to figure out the problem, run those, get the results, and then using all these three pieces of information about the physical space, predicted performance, and the data coming from the probes tell us what is the root cause and what are the next actions.

And the value it brings is that it's going to reduce the time to response for the teams, because when a NOC engineer receives this message, the two, three hours of his job is already done. And also, since it's tied to the digital twin, it tells them where exactly is the problem. So if you need to send people or anything, you know what to do.

I love it. Mm-hmm. I realized we didn't talk...

What is the web address for Eno? Sorry, what's this? The URL for Eno.

ai. Okay. This network observability agentic solution, is that available now?

April 28th. We probably- Oh, okay ... but let's...

Yes. So about almost two weeks. Like that.

Well, a little less. Well, yeah. No, about two weeks.

Yes. Yeah. Well, yeah, two weeks.

Yes. Is there anything, like on the website, we could read about it now and sign up for it, or- Certainly, yeah ... what are-- Yes ...

you can go to our website and there's information about it. And we have intake forms if you're interested to see it, try it. You can definitely fill out those forms.

We'll reach out to you and, yeah, set it up. So Payman, if we're going to go to the website, what else do you suggest we check out there? Yeah.

So our website shows our platform as a whole, our intelligent connectivity platform. It shows that how everything's going to get tied together as one unified piece rather than multiple different tools. It talks about our agents, which we have built many different agents that how they work.

We also have built one of the fastest, most accurate ray tracing models using the latest capabilities of NVIDIA GPUs that helps companies move much faster on their design, get the results almost immediately, versus the current industry, you need to wait sometimes per day to get those simulations run. It's also a fully web-based platform, so you don't need to install anything. You don't need to basically have servers.

And from the get-go, we built it based on collaboration. So it's a multiplayer type of tool that you could have multiple people work on the same design, create videos for each other, have a conversation, upload pictures, give view, edit formats, and even if you want to share it with your customers that they don't have access to our platform, we have ways that they could see and view how things going to look like, all the digital twin design capabilities that the customers want to see, and yeah, and take it from there. I love it.

Mm-hmm. Hey, I want to thank you for coming up here on TechstrongTV. Best of luck with this new solution.

It's interesting times, right? And observability, of course... Look, I think with the increase in the amount of code, the amount of applications, the amount of network traffic that that brings, as we said before, we need humans at the helm.

We can't even be in the loop, where an observability becomes the killer app for governance, security, for just understanding what's going on on our network. So it's going to be an interesting time, Payman. Yeah.

All right. Thanks, Mike. Payman Samadi, CEO, Co-founder of Eno here on TechstrongTV.

Thank you. Keep us posted. We hope to see you soon.

Thank you. All right. We're going to take a break on TechstrongTV.

We'll be back in a moment.