IoT Leaders
IoT Leaders

Episode · 1 year ago

IoT Security: The Largest Problem We Will Face This Decade w/ Peter Doggart

ABOUT THIS EPISODE

In 2021, there will be 46 billion IoT devices connected to our networks.

By 2024, that number is expected to reach 83 billion.

Because each connected device represents a potential breach point, this influx will pose severe operational challenges for businesses.

In this episode, we speak with Peter Doggart, Chief Strategy Officer at Armis, about IoT security and why it is one of the largest cybersecurity threats we will face this decade.

What we talked about:

  • Why the future of cybersecurity is agentless
  • Why the main challenge is asset visibility
  • Steps you can take to improve IoT security

To hear more interviews like this one, subscribe to IoT Leaders on Apple Podcasts, Spotify, or your preferred podcast platform.

Listening on a desktop & can’t see the links? Just search for IoT Leaders in your favorite podcast player.

You're listening to Iot leaders, a podcast from Si that shares real IOT stories from the field about digital transformation, swings and Mrs Lessons Learned in Innovation Strategies that work. In each episode, you'll hear our conversations with top digitization leaders on how iote is changing the world for the better. Let iot leaders be your guide to Iot digital transformation and innovation. Let's get into the show. Welcome to the latest edition of the Iot leaders podcast with me, your host, nickel, the CEO of Iot company Si. IOT leaders is a podcast that attempts to demistify the intriguing, the exciting but often baffling world of Iot, and this week I'm delighted to welcome Peter Doggett, who is the chief strategy officer for Armists, and we're going to talk about all things security in the area of Iot. So, Peter, welcome to the IOT leaders podcast. It's no great to be here, okay, and I think we've got some really cool stuff to talk about. The World of security is full of horror stories and opportunities and intrigue, let me put it put it that way. But first of all let's talk about our mists and must is a pretty hot company. I mean I know that you guys have doubled your valuation in a year. I'm sayingly jealous. You Two Billion Dollar Valuation Company and YOU'RE A pretty unique security company as well, you know, in what you do with the way you do it. So so just give the listen as a little overview of our mists before we before we dive in. Yeah, yeah, would love to and really that this goes back to one of my passions and that is trying to figure out how we go see and secure every single asset on this planet. And fundamentally that that's why arm this was created, was to very modest missions. There was was was try and protect all these new devices and it's the reason I kind of landed. He was again as a passionate mine. It is probably one of the largest cyber security problems we will face the decade and probably in the next decade as well. It's really centered around all these new devices coming on to our networks and fundamentally thinking about their risk to the business and their potential consequences and impact to the business. And we certainly believe that this influx of new devices, these new connected devices, are going to cause and are causing massive operational headaches for almost any business. Yeah, you know, for those of us who are directly involved in this security, into security, cyber security area on a day to day basis, I guess the time we really kind of take the granted. We don't think about it, which will get onto, because that's part of the problem. People don't except upset the bad guys, but the but but put every now and then you get these incredible public horror stories, and I'm thinking just recently, since since the time when we spoke about a week and we go to prepare for this podcast, and now we've had the American pipeline example, of other pipe kbs with almost every day now we're seeing attacks. Yeah, and I guess you're monitor as a company. Do you monitor to those attacks? Can you actually see that the momentum is is increasing? Yeah, this, this impacts almost every single near bact, every industry, from financial to...

...manufacturing to oil and gas affilities, and we are currently tracking over one billion I team devices, or just assets we call them, and over the course of the next four years we have seen and extracklers in our through research, to will likely have about fifty six billion devices by two thousand and twenty five connected, and that's an astonishing amount to divises. There's an astonishing number of potential breach points to to an enterprise. And, and I mentioned that the the beginning, that you do this in a different way. I mean, I know you work semantic for many years. You runder the partner program of the BIZTA functions. But but traditionally, the way, at least of which I understand it, we've tried to catch this ball that's bouncing down the stairs, if you like, is to try and put some software on a device. I mean, you know your your phone, has a bit of software in a corporate version of your phone, your corporate version of your laptop. ARMIS didn't take that approach to you know, and this comes down to a fundamental facts about these devices. And it's also important to understand what's happening. Why are we seeing fifty six billion devices coming out when I works in the first place? Yeah, and past. The reason for this is we are seeing this pervasive nurse because they are super low cost. In the first place, we've got bandwidth glore everywhere. We got low latency. Now, obviously, with the work you're doing neck on the sellilar side. Yeah, Le Be Swab Technologies, and it's just very easy to go and pick up these devices off the shop. Hardware, official software, package it all together. But here's the problem. The numbers of permutations that there are vast and it's impossible to put a security agents call the business onto these devices. It's just practically possible. I think we're tracking on the order of fifteen million different permutations of devices out there. So just think about being a software security company having to design and maintain fifteen million different versions of an agents and somehow working with thousands of manufacturers to get that software onto those devices. It's just it doesn't work and it's a an unsolvable problem. And so so it's so, I guess, just so our listeners understand, I think what you're referred to is they have a certain you have an increasing number of devices, trackers, but you also have different software on those, on those devices and they can be all configured differently as well. That's right. There in a number of permutations that if you're going to their identify a problem, issue a pat a patch, test the patch, this to be the Boutch sot touch. The number of permutations that you have to go through is is it's an exponential problem and you're using a linear if you get trying to put software on it. I guess the way I would describe it is for using linear processes. You know, here's a patch for that device. Well, it doesn't work on this device because it's it's different software or it's configured so differently. We're using a linear solution essentially solved an exponential problem. In other words, you can't keep them. No, and the old way of trying to solve for this problem just doesn't work anymore and in fact, is becoming less a right less relevant. The the ratio of what we call these manage devices like the laptops we're...

...on right now and there we've pretty much solved for that with security agents from people like I can windows and but they're not like IOT devices. They're pretty things, tightly coupled harder and software. It did, as is your phone, but I it devices are not like that. No, no, I'm the only way to solve for that is an agentless approach. It's basically looking at the world from a networking perspective, and we can talk more about that the later that it's almost pioneered this approach about five years ago to take a wholly different approach to security, and that's one of the world we're moving to insecurity has to be agent less in order to scale. We cannot think of this in the old world. So has to be cloud and it has to be agents. So let's double click on this, because we're now getting down to the hub of of of of the approach which, as you say, will ultimately become into the issue of who, who has visibility of the network. But before we get onto that, so you know, the Bullbou sit down says you all the device configurations, the number of devices of being adopted. You trying to put pece software on the computer. Is just can't do it fast enough. In fact it's getting to the point where it's getting less and less relevant every day. Point you make there. So, therefore, paradigm ship is is agentness, which basically case there's an order. Doesn't know what that means. It means you don't put a piece of software on the end use of the device. You actually you actually have central control or central policy deployed automatically to everything that's connected without having to put some code in all of those things, which I guess means that one of the first challenges you've got is identifying all these devices. Is that? Is that where you where are starts? Yeah, absolutely, and I would say a good strategy, a good security Strac to, always starts with visibility. You fundamentally have to understand, whilst on the network, what's in your domain and what's entering your network. And here's the scary thing. Many organizations around the globe do not understand what's on their network, and it's just in worse. I'm sure they have no idea. I'm sure you must find listeners to you ask them advance and then you say, well, actually, we'll put like a snipper on order the technology that you use to identify it them, and the two applete different. Right. There are so many occasions where we walk into a client's environment and we just been a far most in a few minutes and where the client thinks they've got a hundred and fifty thousand assets or IP addresses and whatever. I'm the network. We go know you've got seven hundred thousand and the libel goes off and that's is let a lot of things go up for that. It's quite really that they can be, that they can be that that far about. I was a very interesting case, but many are at least double, forty percent more. Yeah, just goes the point. We've let this trips so much and the technologies have not been employed and the mature it is not cap top with it in security at the pace of the innovation of these devices. So you've got this gap forming that has become so large it's opening up this wound for cyber attacks and, as as you noted, nick earlier, we are seeing countless attacks almost on our daily, weekly basis now,...

...very large ransom bare attacks. You know, they got this picture in my mind as you as think, Peter, of same sort as your house. You say, yeah, I'd locked all the windows and you say to the doors, ever, how many Peo go? I've got twenty five windows and four doors. And you say, well, that's kind of interesting, because I just walk around Your House and you've got a hundred windows and nine dollars. Have I have been really oh, but actually it's a little bit worse, is it? Is that? Is that that there are people adding new windows all the time. So so you it's actually worse than that. Just so there's you can you? Can you conteke analogy one step further? Yeah, there's more windows being added, there's different riots in from different factories of years, and the the guys who are making the locks, well, they're not so good. Yeah, and yes, they're exploitable, and that's what we're seeing is the the numbers are growing vastly, but also the vulnerabilities or the exploits of these devices are increasing. armist over the past numbers of years have released countless vulnerability releases. Ocean to eleven was one we're released about a year and a half ago, and this one was quite staggering. This was just as I was joining the armis organization. A notion to eleven, basically, for those who don't know, is eleven vulnerabilities in some of the largest operating systems that reside in these bote devices, and those devices are things like fire walls, medical devices, operation technology, sensors, manufacturing equipment, all running these operating systems, from the the ex works, nuclear thread acts, Os, the count the list goes on and using the window analogy, it allows the attacker to basically pick that lock within about two seconds and come in. So it's not not only the number of potential ways to get in, but frankly, the security is very, very poor for these devices. So my understaining is so you use agentless, so you'll identify the windows, under doors, the openings. You haven't solved the problem yet. You then profile or the devices, so you then build up sort of databases of this is what you've got, by the way, by type of device, and then I think you can go into you look at the network data, and now we're getting into the solution that we're coming up to. A sell you the devices, which is what this is all about, which is which is cellular. But before we get to that, you actually look at the behavior and the patterns of what's happening on those devices so you can start to build up profiles of it. Our nomally detection behavior, etc. Etc. Yeah, yeah, the system not normal. Yeah, exactly these the system works. If you're familiar with Tesla, and yet they're soft driving software. What were the doing is bracing them to what we've been doing the devices. It's really taking fast quantities of data are in understanding what these devices are doing within context of their environment. So we're seeing tens of nouses, hundreds of thousands, millions of times of a device on a network and understanding how that device is are interoperating, how is communicating, what does it look like? What's normal behavior in a medical context and the manufacturing contexts? And only by guessing this, the some vast quantity of data in a machine learning we have both on supervised and supervised machine learning the back end, can you get to the point at scale where you can, at high efficacy, fingerprint these devices and the as you say, neck down to operating system versions, what protocols are...

...using, what applications they're running, down to look at PLC's in the manufacturing lines. We can actually see inside the cars, the modules and see what message is each module and the commands of those modules. So you can start to gather in an amazing richness of data aging lessly, and the very passage is that of the human. No human could ever do so now. So it's so again, right, but as always, the bull keeps pouncing down the stairs and the attack is get smarter and smarter and there's a whole class of devices that you can't see right there. You know, it's a two companies, armies and SI. We have that a joint announcement. Solution against security solution. About that in a second year. But cellular connected devices in particular you can't currently have visibility. You can't see them that. That's rights to me because they're not connected to the enterprise network. They connected to somebody else's network, the Mokablet. That's a problem. Yeah, that's right, and that the whole nature of Ourmus is. It's passing. We don't actively scan with, as you know, to another agents. So we have to look at the traffic. We have to inspect on the wire, which is good because you get to the source of truth, but you need to have access to that traffic. Yeah, and to your point on the cellular side, I don't know how many mminos there are in the world now. The eight hundred and twenty something according to the GSMA, exactly. So you got a lots of eminos out there and traffic is being dispersed everywhere across loads of different eminos. How the heck to get that data? In a kickings of fashion, it's really real. Don't have the network connection. You just received the data, which may have an attack inside it, but but you don't have you can't, you can't look at the net flow data. You can't analyze the data. So so. So what we just to help the listeners. What we've done, we've announced, is another everyone's open in the sunning. What must do, as those regular bustiness of our podcasts will know, what we do as SI is that we actually run our own private network and we actually federate will Zminos together into a private APN access. Don't use the public internet. We become the single eyed the needle where we managed the connection, we do the billing, we do the device management, we do the support, etc. Etc. But basically we are are with a single Sim we are actually up providing localization, not just roaming, but localization and roaming across those eight hundred twenty networks kilbing on thousand, ninety nine percent global connectivity. And so the cost of the spark that that caused us to talk was you are saying, well, we've got this great business which the Financial Markets Love, not surprisingly given the problem that it's solving. But but actually there's a bigger percentage of devices that are invisible to us, because the edge is growing faster than the number of coup devices. The laptops, printers, PLC's aren't growing anywhere near the rate of IAC devices and we Byre g coming along the you know, people talking about fig is a better version of Wi fi. More and more devices are are going to be cellular connected, which means the percentage of the state that I'm this can see is in danger of declining with our solution, because every device, regardless of which mno is, it's,...

...it's it is connecting to every device, goes through our network where in either needle for all traffic. And so think the the the collaboration that we've announced is that with an interconnect between maybe eye intellect, between armists and sire, the armist functionality is then enabled across cellular networks. And that was the genesis, wasn't it, of why this would be a good idea. We enabled you to see cellular connected devices across any Emino. It's business and it would be good for us, because what we're saying is this is the value of having a private network, not sending stuff over the Internet, which we talked about the EMOs, but a lot of MVA those. So we solve this problem. We aggregated, but they solved it by by actually sending data of the Internet. It's not a they're not managing the network, whereas we run a private network. Therefore we can give you that netflow information and so, therefore you can actually run armies on all the devices that you talked about, plus the devices that are selling that connected through M and opens, and that's the genesis of the corporations. You know, I love this combination and for the very first time and enterprise can now see all the assets. It's on their corporate network, the devices at home in the cloud, on Wi fi and now selling. So, going back to my earlier statements, it's all about visibilities, all about the assets. Where are the assets? Am I getting to a single source of truth? I've understanding where my assets are are and what they're doing, and this combination allows us to go do that and clean up the age old problem of cleaning up the C MDBs, the compigreation, managing databases which power, frankly, operations of any anny good company. And I love what you guys are doing because it's providing racicure connectivity reliably and now we're overlaying the sort of asset visibility and security of the devices and it allows, I think it allows businesses to innovate as well because of the days. What's all about is you can go to one of these angel companies x on a VP or she'll rolling at these ev chargers and now you can have a secure, reliable connectivity constantly and know what's going on with every single charger in their estate and mapping back to their management systems and other systems in their in their own estate. So it really, it truly is in an industry first and so excited about what this business can go and do now with this, yeah, reliable and secure connectivity, which is about to really say a bit of market research in next few weeks and marketing manage or kill me for saying this, but I'll pre announce it because I can about one of the inhibitors, you know, the famous civil billion things are going to connected to about it. A lot of these podcasts, you know, really got to be let and billion things by twenty, twenty or fifty billion and what would inhabits? And there's a ton of them, but number one of every list of security. And so we've been for us, we've been saying, well, that's why, you know, you need a APN based access, encrypted tunnels, private, don't use the Internet. You know, the whole Sim security securities of certificate being stored inside the sin. But actually what this adds is that is that you then can have central policy definition with automatic deployment to the edge, because what you do is bring them policy and the policy can then be dynamically applied to cellular enabled, to cellular connected devices, and now that, if you're a CSOW achieve information security officers say, as well, I'll let these devices be connected to my network, but I dam will need to be sure that my policy. You talk about the configuration management.

You know, you could be configuration policy, compliance policy. which IP addresses, DNS addresses? CAN THEY ACCESS? That my policy. I want to define setting and have alter deployed to the edge. But if you can actually define central policy and having apply to everything that's make it to your network and things that are sending you data that are directly connected to your network, see some point of view you that start saying okay, then, in which case I'll trust the system, because people lose their jobs. They have just lose their jobs. They the company come these guard have a business, I mean pretty quickly, as we know from some of these security hanks that were seeing. So that's a basic premise. And we're going to be roollowing that out with pilots and things. Of course, anyone's listening to the pacifically. They they know which Comans they can contact. But I wanted to just pivot now and just talk about the the the bad actors, the bad guys, because so far what we've been doing is talking about where the good guys are are doing, but the bad guys never asleep, it seems, and it seems to me, Peter the when you hear about really sophisticated security hanks, you think, wow, that was really, really clever. But in the area, and they probably they really are ato and have a lot of knowledge able to get into that system. But actually in the world of Iot it seems like, you know, it's not just the number of windows, it seems like we have an active policy of bringing the windows open and almost inviting people in, because these devices they're not secure to begin with. I mean they're really very unsecure. Only it really is. I mean the we are seeing literally people pick up these devices which have deeper passwords not into them or back years. And you may think, oh no, no, that doesn't happen. Oh No, it does, and it's happening at a lammy rate because it's not at the forefront of people's minds as they are design these these devices. They're at the end of the supply chain, so they're picking up hardware components that have been developed another manufacturer in they Aliti and from Ye from chip manufacturers way back when, and you can get the very simple attacks going. Okay, we're going to try the easy vulnerability of the Remote Code Exploits and and we know there's many, many of the Liberals is out there still the Osians eleven us book about that was a year and a half ago. Guess how many of those devices have been patched? Not many. Three percent were there are the two billion devices. A year and a half, a few year and a half, only three percent have been patched and it's it really infuriates me in a way because we're not taking this seriously enough and it's going to take another colonial attack, it's going to take another Irish hospital attack or a JBS or you name it in the past couple of weeks and we need to have a week of call and we need to get better to put our suppliers and ourselves more accountable for designing these devices more securely in the first place. Only then we're going to get to a better place. But that we've been trying to do this now for any decades, and it's yeah, I was going to say we, those of us have been around a little while. Chae, hold my hand up. We have been safe in this, but the problem just seems to get bigger and bigger and bigger with the explosion of the end points in the network. We just it's a I mean the kind of weird thing is if we think the problems hard now, and it is compared way harder than it was three years ago, which was way harder...

...than five years ago. But in three years time, I mean you know, being a dedicated Iot, can move to as a customers round and the globe, a million devices in the management. I mean we say the customers, you ain't seen nothing yet. I mean we you know, you got the eleven billion, fifty billions stuff that I quoted. But this this thing called massive iote coming and and fight gene. You know massive I it this it's expanding in two different ways. This there's going to be sensors that use some form of cellular connectibility that are going to be disposable. They're going to be disloting imprints, circuit print the battery. You know there's going to be going to be trapping on boxes clothing eventually over wrapper and chop or serning. But but put it. But you know, I've already had a big conversation today a with with a company that's looking at doing proud five gene networks, where you're talking about tens of millions of things in a campus or in a large factory. So another way to the numbers. The the rate of adoption is accelerating. It's not linear, back to the exponential linear. And so, even though, at the very least you can't put if you said one. Are The lessons from this? You know, a, you can never you never know what. If I was summary You, nobody has a clue what what's connected to their network. Is What you've said. Yea. Secondly, you can't, even if you did have a clue, you can't put a piece of software on them because it's really old operating systems and it's it's there's too many derivatives. It's just an edgine. There's not enough money in the world solve that engineering problem because of all the all the permitation, so conoput software, so so see to use an agent less technique, which is what farmist is all about, the other world leaders in that Dan even then there's a whole bunch of advices that you can't see. Therefore, you've got to be able to see the devices that are are really growing exponentially, which other cellular connected devices, which is why you need to go through something like De side, which is the eye of the needle through which all the network traffic flows for all those devices, regardless of the M o. But then what we're saying to it, as we get towards the end of the podcast, is that but even that's not enough because ultimately it's just growing so fast that actually it's almost like we need to do something else that's not tech, just technology. It's almost like we need to raise awareness right the way back to the start of the supply chain, because it seems like we're we're constantly, you know, like the argument on the on the head National Health Services, that you know, not a much money you throw at the NHS, there's never enough, but actually diverted some of it into disease prevention. From an Roy point of view, money pat is spent, but the disease prevention is almost like thinking about how devices are made by people have no consequences of it because they just make devices and sell then and actually building security in a security considerations much more front than it seems to be that we're doing right now, because we're building these these houses, if you like, with the windows open. It's not like, well, thinking about that. They have to do it. No one is really set selling the fact that their devices can't be hacked. It's not the way we're selling things right now. Yeah, there is a slight silver lining, so do want to end on a slight positive notice. Is Not all bad. Like we we do have the tools...

...to help, at least now. But also, I think what we need to do is change organizational and how we think about this problem. You started to mention it in the the the IT teams, the security teams, have been moving slow together. They've really got to collaborate now. They've also got to extend their collaboration to the operation teams, to on the ground of the manufacturing centers and bring them into the fold, and they've got to work with that R D and suppliers as well. Going as far back as possible to fundamentally understand what the risks coming into the organization are. And the goodness is that there are models out there and there's the nest model, there's the CMMC model that's being more adoptive now, and I know things like the federal government too in the US is pushing all the suppliers to get to you must be level three of the security model to do business. So when moving in the right direction, are we're moving quick enough? Oh No, but we certainly need to solve organization how to think about the problem. We need to enhance our maturity of how we think about cybersecurity and I think fundamentally we need to take I think the human out of the loop. So going back to as we engineer these problems, as we engineer the hardware and software, particular software elements, we're always going to because we're flawed as humans, we're always going to put in vulnerabilities and exploits and hears. So if we can, and it's called ship left in the in the world of it here, but if we can shift left and start to remove some of the human elements of doing bad things in the first place. To analogy about prevention on the Healthcare Fund, we can get to a better place, and there are are lots of companies focusing on that part of it as well. Develops, get SEC ops. Call it way you will. It'd be great if we can build a machine that builds these machines. That, guess, the point where that machine is so clever it doesn't have any exploits in there. But then if you start to get into the big AI world of okay, the overloads are going to take over, we will lose control. But you know it's yes, it's a bit's a big old product them. But in the meantime, what we're talking about here is such a order of magnitude improvement of what we're doing, as essentially what we're doing is we're using software techniques to recognize patterns of ustage and normal behavior and identify normalise across millions and tens of millions of devices almost real time. Yep, had you my pass that are connected to somebody else's network, but ultimately they find a way of sending data to you. That itself is mind blowing compared to what you know when you were instamante or when I was in in Cisco, which somebody said you be able to do this one day in the near future, we'd have said, I don't know what you're talking about. It it just being able to do what you've described is mind blowing compared to how we previously pride to solve these problems. That's right. It's nice and down and the CIA, as we speak to us, says we speak to one what will sunderstand what this can do and be a single sort of truth for them and help them see everything that the shoulders. So often it's a yeah, yeah, yeah, you know, competer that. We probably leave it there on that high note and wanted to just thank you. I would add that also for those people who do listen to this podcast, who are often in many cases, and we do get the messages. Considering there in the other stages of IOT projects. I think there's another lesson here, which is that you can add seecurity and peace of mind in terms of secret Central Security Policy, depoint cellular I devices at the edge...

...to one of your checklist of things that you can look at, which will probably mean that your project is likely to get a proved because a lot of projects don't get approved because they get blocked a security concern. So I think we've also made an IOT adoption to be least, taken away one of the barriers or one of the main pushbacks that we get on projects, which is no, it's not secure. I know this is a business outcome, but it's not secure, so I'm not going to do it. So I think that's a big improvement as well. So we'll leave it there. One last question to you, though. If anybody listening to this wants to get in touch with you, what's the best way for them to send you a message or ask you a question? Maybe is linkedin or what's the best way they can? You can certainly look at me Linkedin, Peter Dogret and you can find me. I am the only Peter Dogritt and cyber so that's eusly. Okay, that's the OGG Aart. That's right. Yeah, it's not easy. Great. Well, thank you, Pezza. Thank you again for your time. Very, very interesting. We haven't done a deep down on security before, so that was really, really good. I suspect that in a year's time we could probably repeat this and do a retrospective look back at some real case of these and stories, and then that would be really interesting to see what people are doing with this. So that's probably one of the goals of our alliance. But in the meantime, thank you to you, the listeners, for listening that this was the, as you know, the IOT leaders podcast. If you want to send any feedback, you sent to me nickel and on linkedin or an email to IOT leaders at Si. That's the see yecom. In the meantime, thanks again. Look forward to the next podcast and, as they say, maybe an updates on this very exciting area where we should we know there's a Latmo to become in a few months time. Assist solutions are very exciting solutions starts rolling out in the market place. In the meantime. Thanks again, Peter. Thanks to you, the listeners, for listening and this was the IOT leaders podcast. Thanks very much and take care. Thank you. Thanks for tuning in to iote leaders, a podcast brought to you by SI. Our team delivers innovative Global Iot cellular connectivity solutions that just work, helping our customers deploy differentiated experiences and disrupt their markets. Learn more at SICOM. You've been listening to Iot leaders, featuring digitization leadership on the front lines of Iot. Our Vision for this podcast is to be your guide to Iot and digital disruption, helping you to fought the right route to success. We hope today's lessons, stories, strategies and insights have changed your vision of Iot. Let us know how we're doing by subscribing, rating, reviewing and recommending us. Thanks for listening. Until next time,.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (30)