Simbian.ai’s Ambuj Kumar Explores AI’s Dual Role in Cybersecurity

This is Textron tv. Hi everyone. Welcome back here to Techstrong tv.

Got a first time guest here on Textron with us. ai. And his name is Amou Kumar Amou.

Welcome to Tech Drunk tv. It's great to have you on here. Pleasure to be here, Alan, looking forward, Great Amou.

We're gonna talk about AI and cybersecurity, but before we get to that, I always like to kind of let our audience know who they're talking to, right? And well, it's not me who's gonna let him know it's you, who's gonna let him know. Why don't you, if you don't mind, give us a little bit of kind of your history and your, you know, your own path here today.

Awesome. Um, so I grew up in India in, um, a rural, uh, place. Um, I came to US about 20 years ago, uh, to work for a small company back then called Nvidia.

Mm-Hmm. Which, of course is now running the world. Yeah.

And, um, so I was, uh, with Nvidia for eight years, built multiple generations of GPUs, worked very closely with Jensen Wong, um, the CEO. And, uh, later on, um, I, uh, migrated to doing cybersecurity. And, uh, I ran a company called Fornix, um, um, one of leading data security companies.

Now I'm ion symbian. Um, so it's, uh, funny that I have come full circle, uh, from building AI to now using AI for cybersecurity. Well, you know what's interesting is, is you're right, right?

For for how long was Nvidia not an afterthought, but certainly not a leading shape player, right? And, and if someone would've come to you then and said, you know, NVIDIA's gonna do great one day, let me tell you, you know, what, what role or what, what's the market that's really gonna set them on fire? It would've been, well, you know, the more games people play, the more graphical, you know, GPUs we need.

And, and, and that's where they make their money. Funny things happen in real life though, right now AI happens and Nvidia is what, a two or 3 trillion or $2 trillion company, I believe, right? Maybe more than two now.

Uh, depends on the market any given day. And then you go off and do this cybersecurity thing, right? Which is a hot market.

And then circle back. And here we are now, where kind of your world's come together, AI and cybersecurity, right? So that, that's a fantastic story.

And it's, it's real life. This is what happens in tech. You never know what, what that next hot tech thing is gonna be.

So good for you. Let's hear a little bit about Symbian, if you don't mind though, what, you know, give us the background there. Yeah, so Symbian is a cybersecurity startup.

Startup and startup. If you look at security, um, there are of course so many products, so many things, so many issues, and, uh, it's never ending. Um, you know, cycle of pain.

And now hackers are going after hospitals, critical infrastructure. So it's not about just losing our money. Cybersecurity is today about, you know, um, just being alive.

And so I thought very deeply on what is the root cause of all these problems. And I came to realization that security is like health, right? I mean, if you want to be healthy, you gotta exercise, you need to make sure that you don't get in accident.

Um, you need to take good care of your cardio and weight and nutrition. So it's a bunch of different things all working together. Same thing is true in cybersecurity.

It does not matter if your endpoint is fully secure, if your people are clicking on bad emails, or it does not matter if your data is secure, but your network is not. So you need to have whole bunch of things work together in harmony. Um, but, um, the attacks that happen in the defenses that are put, they are on various specific topics.

So it creates this, you know, conundrum where organization needs to have 45 different tools, um, work together. And so they need, you know, 45 different people or teams, uh, who are specialized on those things. And, um, um, you know, there is, there is of course in a stat that 3 million security jobs are unfulfilled because it's so hard to get talent.

Um, so I realized that rather than specializing in each of these topics and hoping that organizations will have talent to cover those, what if I could build an AI engine that understood all of the security concerns, all of the security products, uh, all of the security issues, um, and, uh, you know, like that could control your security. So we call it fully autonomous security, uh, early days, but that's what Simeon is building. Absolutely.

And it is early days, right? I, I think one of the things we, we, you know, I've been in, in technology 30 plus years, security, 20 plus years, um, 25 years. I've never seen a technology wave so fast.

So, you know, become so big as, as, as we've seen in AI over the last year, year and few months now, right? But we, we, and it's, it's become so fast and it's become such a part of how our consciousness of how we think about how things are gonna work, that we, we forget, it's really immature right now. A lot of, a lot of what we see is the promise of this not not delivered today.

We see it when we talk about AI writing code. Yeah. It writes code today.

Is it perfect? No. Is it it better than a developer?

Probably not. But can you see with what it's doing, how it can quickly leapfrog? Yes.

Is it, is it a big help to cybersecurity today? Maybe not. 'cause we're still sort of figuring out, you know, how this interacts, but can you easily see where it becomes a 10 x you know, uh, multiplier going forward?

Yes. And I think this is, this is, you know, the conundrum. We, we, we tend to focus on the promise and forget a little bit about what's possible today.

So let, let's talk about Symbian, but as part of this bigger discussion of what's doable today, amou, and then what do you see being doable in the near term future? You know, that is so correct that even, even outside security people talk about a GI, artificial general intelligence and idea is that can we build an AI that will do everything for us? And people are understanding that, yeah, maybe it's possible, but it'll take, you know, 10, 10 years.

It's not, it's not coming, you know, this year it's not coming, you know, next year. Um, same thing is true in security, right? Um, like we cannot build an AI engine that takes care of everything.

Um, in, in, you know, this year, next year we cannot because technology is not, um, you know, available. Um, but what we can do is we can use symbian today to do various specific things. So for example, let's say, um, you, you want to write email to somebody of notifying that his, something happened.

Now we have seen, you know, just like Chad, GPT can write poetry, um, LLMs are good at communication. Um, you want to summarize bunch of steps you took to resolve a ticket. I mean, LLMs are good at that.

So whenever there is language skills you can use, um, ai, AI for that. Um, from symbian perspective, we are right now focused on helping, um, security operations team. So let's say that, you know, you are responsible for monitoring firewall and something bad happened.

You want to understand what happened. Maybe you want to pull information from various tools, put it together. Um, you are a SOC analyst and you are just inundated with, you know, operating 20 different tools.

So we can take natural language command from you and go operate those tools for you. So those are things that can easily be done today. Absolutely.

Um, so let, let's specifically on symbion then, where, where is the offering, if you kind of get what I'm saying? Yeah, yeah. So, um, our product is called Security Accelerator.

And, um, idea is that, um, security is is never a, you know, done deal. It's, it's always a journey. So our product will help you get better, uh, you know, they're faster.

And, um, so we are working with a, um, bunch of customers and partners right now. Uh, we have a few design partners. We have channel partners, so early days, uh, but, um, customers are using us to, uh, centrally control their SOC team.

So, for example, one SOC analyst has best practice of doing something. Somebody else has something different. You want to find, um, you know, like best of everybody's, um, um, you know, methods.

You can, you can do that. Um, you, you want, you got an alert, you want to enrich it with, uh, you know, various threat advisories, threat intelligence, um, you want to look up internal systems. Um, so rather than you going and, you know, clicking 50 different products and 50 different buttons, um, Symbian can auto automatically go and fetch all that information from various sources based on what you are doing.

So if it's a phishing attack and, uh, you know, like will introduce it one different way. If it's a phishing attack of a particular type where, um, you know, sender has done some damage in past, we'll enrich it differently and pull all that information and, and give it to you. Excellent.

And this, this is like sort of a sa it's a SAS offering. Yeah, yeah, Yeah. Um, if I ask you to pick out your crystal ball now though, and tell me what is symbian gonna offer a year from now, or 18 months from now?

How does this offering mature, grow, get better? Yeah, I mean, so, so this, this brings back, you know, where we wanna go. And I firmly believe, Alan, that not only AI can help us do security better, but that is the only way of surviving in security world.

Um, because attackers are going to use like sub chat, GPT or purpose-built LMS to automate, send emails, to do research on individual employees, figure out, you know, like where they are susceptible to do recon on your environment. Uh, find out that you are using one software that is vulnerable, can write malware that is precis precision, you know, developed for you. Um, and so if they can do an attack that cost them 1 cent, and you are on the defender side going to spend, you know, a hundred dollars with people to respond to those attacks, it's not gonna scale.

So in the end, we will have AI doing everything. Um, and the way I position is that it's fully autonomous security where humans control which way we wanna go. And AI is there to, you know, move the motors, um, per your, per your command.

So 12 months from now, 18 months from now, we will see symbian take higher level directions from you. So, for example, you can describe, this is how I do identity management. And, uh, Symbian will understand that and, and, you know, automatically go and do certain things from you.

So today, um, it, it can take for example, example, you know, like a, a ticket as I mentioned and figure out how to enrich it tomorrow it can take a ticket and figure out how to resolve it, you know, day after it can understand that the reason we are getting so many tickets is because there is this one underlying vulnerability in your system, so we can go resolve it. So it's a hierarchy of abstractions and uh, sure. You know, we'll, we'll continuously go up.

Baush. I wanna bring up another area that I, and I, you know, you're not the only person in cyber that I talked to AI about. We, we, you know, it's a very popular topic.

Is what symbian doing here strictly generative ai? Or are you using sort of that ML AI in conjunction with it? Right?

So we're recognizing patterns, we're, we're learning, if you will, based upon what has taken, and then we recognize those patterns we learn, and then sort of the generative AI part is here's what we do to be better going forward. Yeah, I mean, uh, again, great question. So, um, we use various different tools because it frankly doesn't matter what you are using to give your solution, it, it needs to work.

So we use generative ai, we use traditional ai, but we also use lots of, you know, good old rule-based software. Sure. Because guess what, you know, like sometimes rule-based software, or most of the time rather rule-based software is better.

It's faster, it's, you know, precise, it's accurate, you know, what you are getting. It's just that there has been gap, and that gap can now be filled with LLM. So, um, we use combination of all three.

Agreed. I, I, I could see that, um, Abu you know, I don't know, did we mention the website or we didn't even tell people about that. That's a good idea, but people wanna get more information about Symbian.

Where'd to go? Yeah, www dot symbian do ai Symbian is spelled S-I-M-B-I-A-N. And one of the jokes I mentioned is that early days you have to spell your website and your name as your company gets more popular, you know, people will understand what we're talking about.

So for people listening, it's definitely early stage company. Uh, love your support at this point. And one day, you know, I make everybody proud.

I'm sure, I hope you will. Hey, you could be the next Nvidia, uh, Ambu Kumar co-founder, CEO of Symbian ai. That's S-I-M-B-I-A-N, in case you don't know how to spell it yet.

Dot ai, uh, early stage company, AI and cybersecurity. Check them out. We're here on Text Drunk tv.

We'll be back in just a minute.