Waves with Wireless Nerd
Join me for a weekly look into what's making waves in tech and the wireless industry! What's new? What's now? What's next?
Waves with Wireless Nerd
Revolutionizing Network Management: Meter AI Innovates Monitoring, Control & Configuration with Command
Get ready to unlock the future of networking technology with our guest, Anil Varanasi, CEO of Meter, as we promise to equip you with insights into their groundbreaking advancements in AI and wireless tech. Join me, Drew Lentz, the Wireless Nerd, along with special guest, fellow nerd, and friend Keith Parsons, as we explore how Meter is revolutionizing the networking landscape with their introduction of Command - a generative UI product that enables IT and Networking teams to query their network, take action, and create custom software fit to their specific needs – all in natural language. This AI powered network monitoring, administration, and configuration tool is powered by an AI engine that is trained and constantly updated and evolving based on YOUR network. To kick things off, discover how Meter simplifies the daunting tasks of ISP management and billing, offering users a seamless and efficient internet solution likened to a Kayak or Expedia for ISPs. Then get ready to interact with your network in ways previously unimaginable.
Ever wondered why IT infrastructure upgrades often stall due to emotional attachments and cognitive biases? We'll unravel the concept of sunk cost and share a strategic buyback program to ease the transition pains. Learn from our personal experiences and the expertise of Meter’s team as we emphasize the value of innovation and strategic planning in overcoming these hurdles - and discuss the buyback program that exists at Meter to help move you into the future be getting rid of your dead tech weight. Witness the practical application of Meter’s products, which highlight their impressive capabilities and the importance of hands-on problem-solving in refining technological solutions.
But that’s not all! Get an EXCLUSIVE live demonstration of how Command streamlines network management, making complex tasks a breeze with a user-centric approach. See how real-time network monitoring becomes as simple as typing a question, thanks to advanced AI and automation. Finally, we delve into the future of networking engineering and Meter’s innovative plans to maintain their mission of providing the fastest, most secure networks. Don’t miss this episode filled with insightful discussions and exciting sneak peeks into the future of networking technology. For more information, visit Meter’s website or follow them on LinkedIn - meter.com or https://www.linkedin.com/company/meter-com/
Experience Command at https://command.meter.com/
Hey everybody, what's up? It's Drew Lentz, the Wireless Nerd, and today we have a really special and fun episode. There's been lots of talk about AI and what's going on in the networking space and the wireless space, and you know all the commotion that's going on. And a couple of weeks back I guess not weeks, a couple of months back I got a chance to go visit the office at Meter in San Francisco and I had a fantastic time, got to meet the whole team, but, more importantly, I got to see a sneak peek of some really interesting stuff and I've talked a little bit about it without giving away any details.
Speaker 1:But today we've got Mr Anil on the line. He's going to show us what he showed me back then, except it's going to be new and polished and fun. And I've got a special guest. Mr Keith Parsons is on with us right now. So Keith and I are going to have a good time and we're going to look at what Neil's got to show us. Oh, it's going to be great. Keith, I'm assuming you have seen some of this also.
Speaker 2:I have. My job here is just to be color commentary. You're the main broadcaster. Sports coach.
Speaker 1:Oh yeah, I'm just going to disappear from the screen and I just want to kneel to show us what's going on, because and I, I, agree with your your excitement.
Speaker 1:When I saw it, I was also very excited, so I'm looking forward to seeing it again I, I, I felt like and I think I shared this when I was there in san francisco I felt like we had gone back to like a dot-com era of innovation. I felt like I'd walked into this office, you know, and it's the old school San Francisco startup, like you know all the nitty gritty and you know, peter the chef was there and everyone was chilling in the break room and it was such a good time and it wasn't. I mean, there was that feeling. But then when we started to look at the innovation and you were so cool man and I do I always appreciate that hospitality where you were just like just show them whatever, whatever you want and you know, true to my word, you know we didn't, we didn't sign anything at the time, which was really cool, but you didn't say anything and I think that that was more fun is understanding that there was something coming and you know you showed, you showed me the whole suite of hardware and software.
Speaker 1:So if you could catch us up real quick, because there may be some people listening or watching that they don't know who you are and what you're doing. So if you could just catch us up on on, on who you are and where you're at now, and then I mean oh, then then show the goods, man. So what you got, you got Anil. Who are you, sir, and what is your company? What do you do?
Speaker 3:Well, thank you first for having me. I think you and Keith taking time is amazing. As I was saying before, we hit record wireless royalty, so I'm really happy to be here. I think you both know a little bit, but maybe for listeners, I work at a company called Meter. I'm one of the people that helped start it. I'm an engineer here, and what we do is make networking much better.
Speaker 3:We think networking hardware and software has gotten the love it deserves over the last decade, and that's essentially all we spent the last decade doing is how do we make networking hardware and software fast, secure, reliable?
Speaker 3:And since this last decade we've been doing it as a service, and for us what networking as a service really means is combining great products, great service, together without compromising any of the control and visibility customers not only need but actually honestly should demand from vendors they should have. Somehow it's become that if this thing comes as a service, it's a black box. I can't touch it, I don't know what it is, or it might not be on par with the legacy vendors as far as feature sets go. We don't believe in that at all, so we've been really investing in that. So Meter makes it easy for anybody to get really great internet infrastructure that's both customers and partners across ISPs, routing, switching, wireless security, dns security, sd-wan VPN all in a single platform that's tied across hardware that we design, software that we build and networks that we help make sure are fast, secure, reliable.
Speaker 1:So, and starting at the beginning of that, one of the coolest tools is the ability to choose your ISP and choose your connection and choose your mechanism on how you get to the internet, and that's a software service that you'll have. Can you tell me just a little bit about that, sure.
Speaker 3:I think. Interestingly, that actually was a pain point my brother and I had about 15 years ago. We were running another business and every time we moved offices it was frustrating to us that we could buy everything on the Internet except Internet itself. We just didn't understand why, and that's where the kind of core frustration came from, and we spent about four or five years actually wrangling different APIs, manual data processes, and today it's a product called Connect that's been around for a while and growing really fast, and it's a beloved product amongst customers.
Speaker 3:Honestly, people really love it and the simplest way to think about it is Kayak or Expedia for ISPs, including ISP management and billing, all in a single space. So you just go to Connect and you say I am moving to 123 Main Street. We go, bring back all the ISPs that are available and we partner with not just the largest ones but also the small, local ones as well, and we'll show pricing and availability all in a single place. And not just that you can just buy it and get a contract that's landed in your inbox and all that information is again fully integrated into the entire network. So when you're looking at your security appliances, you know what ISP, you know what speeds all of it. But Connect is Kayak or Expedia and billing and management in a single place and people can go use it. Today it's at metercom, m-e-t-e-rcom slash connect and just go type in an address and go use it.
Speaker 1:That's awesome and there's no strings attached to that. You can go, punch it in and you can use it today. It's such a fun product. I've had a blast showing it off to people and just take a look at this. So that starts at the very basic right and then starting to move up the stack of what meter offers. You know there's, from a comparative perspective, there's people that are moving into the space and whether it's an existing managed service provider or it's an equipment manufacturer that offers these managed services or network as a service.
Speaker 1:I know that you all do things a little bit differently and you mentioned that where you, where you don't just want to say this is what you, what you get, is what you get. It is what it is like the worst, you know the worst four words you can have there Um, but you give people the option to. This is how it should be. But if you want to make tweaks and changes and whatever, you have the ability to do that. So, moving up the stack if we take our internet service provider, we plug into your what, your security appliance, your router, your firewall. What's the first step after that?
Speaker 3:It plugs into our security appliance that has full, stateful firewall, high availability monitoring, including host monitoring, configuration of DNS, dhcp all built in there directly. From there it goes into our switching platform, and both the security appliances and switches are not just 1 gig but 10 gig as well. Awesome, and the same thing on the switches, too, being able to change anything from broadcast storms to where ingress and ingress are going, power on particular switch ports, all full control that connects to our wireless access points. Our wireless access points are indoor and outdoor, again fully configurable, down to captive portals that are built in radius profiles, all of it. And all of this is with our dashboard today, with an end-to-end platform where the ISPs in there Connecto. We talked about then security appliances, switches, wireless access points, and then SD-WAN, dns, security and VPN all in a single platform.
Speaker 1:That's awesome, and that end-to-end infrastructure is key in kind of your next innovations, right? Being able to control the packet from start to finish, all the way from the ISP, all the way out to the wireless, you know, to the person connected to the Wi-Fi, from not just a routing perspective but a security perspective and everything in between. And I think that there's something to be said for that. When you look at different infrastructure models and you look at the way that people, what their networks are comprised of, they have all these different things that are out there. Being able to control that packet to start to finish is important, but it's also to make sure that it's cost effective.
Speaker 1:When you have the ability to do that, you don't want to pay. You don't want to get into a model where it's like, okay, well, I have this router firewall. To get into a model where it's like, okay, well, I have this router firewall, I've got this security appliance, so I might as well get the AP from this manufacturer, but then you're paying out the nose for that AP or I've got to get a service contract on that switch in order to maintain it. Your structure for not just your data model but the pricing model as well, lends itself to being able to easily get invested in an end-to-end meter network. Is that a fair assumption?
Speaker 3:Yeah, absolutely, and we see this all the time. That includes all of our customers and partners and their customers that might be dealing with environments where there's a bunch of legacy hardware. On top of everything you said, we also have a buyback program where we will buy out existing legacy hardware, so there's no sunk cost for our customers and switching is easier into this platform. That's end to end. And Drew, you know this, and Keith as well where we didn't come out and say anything for a long time about who we are.
Speaker 1:I mean only a decade right.
Speaker 3:Because it's such a mature industry.
Speaker 3:Right, networking is one of the oldest industries in all of technology and if you go look at market caps of companies in this, it's second or third largest in the entire world and it's one of the most core underpinnings of almost every business.
Speaker 3:So when we come out and show products, we didn't want it to just be a point solution or we didn't want it to be something like oh cool, that's a nice little startup, you got there, but it can't be X, y and Z right when people see Meter. We want it to be a substantial new way of doing things that's entirely end-to-end and it doesn't matter if you're somebody that's technical or you're incredibly technical like you and Keith, and really complex environments with hundreds of locations and manufacturing or schools, whatever. We wanted you to have the entire control, the visibility that you would expect and should demand from enterprise networks, that you would expect and should demand from enterprise networks, and that came across from the hardware that we built, our architecture, our software, our services and then pricing all together so that, as people come onto the platform which is why we're seeing such growth is because we think it's a mature platform now rather than something that's just on the margin yep, now, keith, you've seen this, and you, you've seen a lot of hardware and software.
Speaker 1:Man, I mean, it's it, that's the beauty of being in the industry saying I'm old, I get it no, no, you.
Speaker 1:You know, I was like I was watching a trailer for a brand new 2024 EGA-style game and my wife looks over at me and she goes what are you watching? I was like it's like Crystal Palace, but for 2024. And she says what are you talking about? Or Crystal Caverns? She says, what are you talking about? I said, never mind, we're not going to talk about Apogee software and Duke Nukem and any of that.
Speaker 1:But I know you know what I'm talking about just by saying it, but that, with that in mind, you've, you've seen iteration after iteration, what I mean, what sticks out to you so far, before you talk about the good stuff, what sticks out to you so far about the meters stuff? Because it's, it's there's. I mean, this isn't like a fanboy thing. I definitely. I am a fan of what you're doing, but I I am critical also, and I know keith is critical too. So there's, there's things out there like, okay, this is a big plus and this is a big minus, aside from the minuses right, let's leave those for another conversation what are the big pluses? You see, with Meter? I mean, you've done this before.
Speaker 2:Anything stick out way back machine. The first time I took a accounting 201 or something, I failed my accounting class because I did not believe in sunk cost. The instructor taught it and I had just bought a truck and he used me as an example and said you know, that is that sunk cost. It doesn't matter. On the next decision, your next decision is from now moving forward.
Speaker 1:That's right.
Speaker 2:I could not get over it. I wrote on the final a big three pages about how wrong he was. Next semester taking the course again, of course, the professor a different professor of course calls my name out loud and makes me stand up in class and he goes you, you're the sunk cost guy, like they'd obviously talked about it. And he just goes, just come talk to me and I go into his, into his office, and he's like you think it's religion and you think sunk cost? Is this this thing? You're? You just don't get it religion. And you think sunk cost is this this thing? You're?
Speaker 2:you just don't get it so just if you want to pass a class, just give them my answer and I'm okay. Well, I think a lot of people are are stuck when they're looking at some of these, uh, managed services that they believe in that religious belief.
Speaker 2:But but I have all this money. It's it's all sunk cost. Some cost is not a real thing, it's an accounting thing and there's a lot of emotion tied to it. So I knew when you said you have a way to do a buyback. Really you're just lowering that threshold of someone's pain point because they think it's really bad. And you know I'm sure you're crying across this, true, drew, I have people who are still using 802 live11b because they believe that they have enough devices to justify that. Until you, I've got a spreadsheet where I put in all their numbers and I show them. And the cost to upgrade your infrastructure to support B and new clients at the same time is way higher than anything you'd ever have to pay to upgrade your clients, but it just hurts.
Speaker 2:So, anil, thanks for even mentioning that and having a program that will help people, because that's a stumbling block and I think those are when people are thinking about companies like Meter. Where am I giving up all the control? Well, kind of. Am I getting in bed with only a single vendor? Well, you probably are already. So there's a lot of those little bumps along the road. But having a mature product and I've just got to congratulate you. You stayed stealth for a decade yeah, that takes and I didn't mean that as a joke. I meant you had the ability to hold back and stay. It's like a charging horse wants to go someplace and you're like no, this is our plan, this is where we roll out and we're going to be ready, and so that took some maturity to pull that off. So congrats.
Speaker 1:That's awesome.
Speaker 3:Yeah, so okay. We would rather win the war than the battle, oh man all right.
Speaker 1:Well, so, so you can. You've got access to the isp, you've got access to what it plugs into, you've got you've got a network that's running in the end meter and it's it's doing all the things it's supposed to do. I you know, your team is as keith mentioned. I mean, your team is top-notch, not only for being able to keep a secret, but for where they came from and what they've been able to do and the innovation that's come based on the experience of the people on your team and what they've been through. Their battle scars have turned into your benefit. You know, and and it's such a good team of people that are there and you know all you got to do is look at the website and see who's behind you all, who your employees are, who your investors are, and you go. Man, these are people who've seen a lot of it. This is a great way to innovate, and so it doesn't stop with what we're talking about. It doesn't stop with the circuits and doesn't stop with the internet access and the ability to control all of it and top-notch hardware in a great package at a great price.
Speaker 1:What you showed me. I had problems understanding how I would use it. It was so compelling that I couldn't even think about a way to use what you had just given me the ability to do, and I think that I think that that was my first and I think I told you that that was like my first concern is, like, now that you have all the power to do anything, this is like the buffet mentality. Well, now that you want everything, like where do you start? And so I hope you have some examples of where to start for feeble-minded people like me, because it's so much and I would love for you to share that and it's been a while. Can you talk a little bit about what brought it on? You don't have to talk about when you broke it, you know, a month or two ago, but you know I love it.
Speaker 3:I broke it last night too, so it's okay.
Speaker 1:I love it, but that's you know. The whole thing that I think is beautiful is that you're not someone who's sitting back going do this, do that, do this, do that. You're actually breaking it and fixing it and breaking it, and other people are fixing it and they're breaking it and you're fixing it to a point where you are so hands on with this because you want it to solve the problem. So, if you can talk a little bit about why you thought about doing what you did and what was the path to get there, and what in the world are we talking about?
Speaker 3:Yeah, Before I jump in, I do want to like nudge. One thing here is I tend to get on a credit all the time everywhere I go, but I'm just literally a figurehead, honestly, and I get to do the podcast with you guys and you guys come over, I'll hang out with you and it might seem like I'm the one doing things. But, like you mentioned, I think what we've been really fortunate with is having people that really care about building great products and whatever the analogy of standing on the shoulders of giants. It just turns out for me the giants are my team, so I get to like play with them and learn from them, and so whatever credit we kind of get, I want to make sure it's for the company. You know there's there's this kind of saying of you know you should play for the front of the Jersey, not the back of it. I really believe in that. So I'm yes, I'm deeply involved and I care about products.
Speaker 1:You say that by the way, and just so anyone that's listening or watching, there are two Sarahs on the line that are hiding behind the scenes.
Speaker 1:I'm sure they have hot pokers to stick a nail with if he says anything he's not supposed to.
Speaker 1:And I didn't realize, you know, just on that note before you jump into it when I was there, when I was at the office, it was just again, all I could think of was that whole early dot-com stuff, not in a bad way, in a good way that young, innovative fun where everyone is rowing at the same time and you walk into your office and it wasn't just the aesthetic of it. And it wasn't just the aesthetic of it, it was the feeling of the employees and their excitement in sharing things and how everyone just completed each other's sentences as we moved across each product and platform. And not to take credit away from you, but really you are not the only person there and your team is a phenomenal team. And if anyone has had the opportunity to come by or visit with some of the team at Meter, you'll know that each one of them can carry their own weight massively. But there's also something to be said for someone who can congeal all of that and make it something. So you and your brother are doing some incredible stuff.
Speaker 3:Yeah, we both really deeply care about products. We care about what we put our name on and at the end of the day, I hope you know we contribute something to the industry and we try to stay close to everything that's going on. And to your question on what's the problems we're trying to solve, you know I studied networking in college and you know my kind of real deep journey into networking started about 20 years ago now. The awesome thing was you came into the lab and you were instantly given this power, which was like the command line, and sit down and then they're like type this and you kind of learn, you know this command and that command and you felt really powerful because there's this like humming big machine in front of you in the lab and you're like I can control this thing and as a young engineer, that's like really cool.
Speaker 3:Yeah, I mean you get really excited by it. But, as I was kind of going through learning, command lines are complicated because you have to learn different commands. You have to remember them. Are complicated because you have to learn different commands, you have to remember them and then, as you go from vendor to vendor and platform to platform, you have to learn the shibboleths of each of those on what it takes to kind of get it to do how you want.
Speaker 1:That's right, and some of the people you know it's. I think about hpe back in the day, where in in cisco, ios, everything was one way and then in hpe it was completely backwards and then Xtreme did something different and then each person had their own take on it and you had to learn, like you said, those very small, specific things about which way to flip the command in order to get it done. You're totally right, yeah.
Speaker 3:And the other thing that was great about a lab back then is you had people peering over your shoulder that can help you or say that's wrong, here's how you should do it. And that was great too, because at the end of the day, you're trying to learn and you're trying to build better infrastructure and support good networks. And then in the last decade, decade and a half, one of the things that's amazing that's happened is dashboards got built, pulling real-time data into web apps, and that was also fantastic. You didn't have to go learn all of these different commands for every different platform and every vendor, but also even forget that even if you had a single vendor, there's so many commands to learn for all the different things you want to do. Some you might be doing once a month, some might be doing every once in a while, some you're doing at the nick of time when you're trying to solve a problem or fix an issue, and you can't remember all of them, even if you're a single vendor platform.
Speaker 2:There's nothing more frustrating than that nick of time scenario when the only thing you can do is hit the tab and try to learn, like right now. That's just in time fixing and you don't know the command, but you can kind of figure it out on the fly by just typing in and fill it in for me something. Tab question mark.
Speaker 3:Yeah, that's exactly right. And then the other thing with dashboards is like okay, you want to make a feature request and we all know how companies work, right, you put in a feature request, it goes into product teams, and it goes into design teams and engineering teams, then QA teams and then full circle back, and in good situations that might be days or months, but realistically it takes like three to six months at least before you can get back what you want.
Speaker 1:Well, depending on what the you know, how much money people are willing to spend on it and how big the account is, or if you can get all your friends to check it and say yes, I want it too.
Speaker 3:Yeah, do a community drive for product development. But we've been really thinking about is that, especially as software moved up the stack right it went from assembly to then operating systems and local applications and browsers and applications in the browser We've really felt software has gotten very rigid, and rigid means is that you can't make it your own. And the other thing I've always loved about technology, other than just networking as a whole, is everything from HyperCard to Visual Basic to. All of these were incredible because the promise of software was here you go, young man, you can make it whatever you want.
Speaker 1:That's right. Absolute HyperCard he says it's great and that was the promise right.
Speaker 3:So what we're really thinking about is can we make software soft again, really make it be moldable and malleable around the user, while taking the best parts of the command line, best parts of dashboards into a single place? And that's what we're kind of announcing now is the product is called command and we hope it's the best of that, and what we hope even more is the future of software in general, that's, around the user yeah, and that's I think you know, and that's a tall order, right?
Speaker 1:because when you start to think about the commands and and the the programming if you will necessary to configure router, to configure bgp, to you know, uh, install a security appliance or to set up a wireless network, that's tied to an acl, that's tied to a specific port, that's tied to a profile, that's tied to a vlan, on and on and on. There is a lot of complexity there and a lot of the devil there is in the details of how you do it. One tiny mistype, one misconfiguration can cost you a whole lot of stuff. Look at what's happened recently with AT&T, you know, or you know any of these big organizations where something simple has caused an issue. Something simple has caused an issue and in that is like you mentioned at the beginning, it's not just understanding what you need to do, but how you need to do it and being able to do that across multiple products, even in the same family.
Speaker 1:I ran into an issue two days ago where I was configuring a set of VLANs on an access point and I was configuring them on a switch, but I forgot to configure the VLANs to be passed through on the security appliance and I couldn't get on for some reason. It was like wow, this is still a thing in 2024. What have you done with command to alleviate some of that hassle? And what is that? I don't know if you have something you can show us or not, but what does that look like from an I, from your perspective of what? What do you? I guess we've established what you're trying to solve. How are you solving that? What does command do?
Speaker 3:yeah, uh, I think it's better showing rather than telling, and then what maybe the three of us can do is, uh, play subtitles for the folks that are just gonna listen, rather than also that sounds good to me and neil is sitting behind a white brick background.
Speaker 1:Currently has on a tan polo shirt keith is looking intently through his black framed glasses. Oh, spectacles would have been a better word there anyway.
Speaker 3:Um, so this is command um, but I can start there through if you want to go first, or Keith.
Speaker 1:This is the best. This is where you screenshot and you go this is command, and it's just a blank canvas, which is awesome and terrifying at the same time. I mean, so what can I? It does have two sides.
Speaker 2:It does have two sides. It does have two sides. That has command and then another side and it looks like it's going to have. You know, do this on one side and we'll show you what it works on the other.
Speaker 1:It just says what would you like to see? Well, what would we like to see? I mean, yeah, let me start off.
Speaker 3:So the challenges we're trying to solve is exactly that right. You don't have to remember commands and you want to have software that's yours. That's the challenge we're trying to solve. Underneath is huge amounts of architecture and infrastructure investments we've made over the years. Where is that? That's across operating systems, that's across our firmware, that's across how we send data back, our backend and our APIs. We made these investments the last few years that put us in a position to build something that's actually valuable. But without any of that, from our hardware on up, what I'm about to show you, and the speed of it and the accuracy of it, just isn't possible.
Speaker 3:So you mentioned ISPs just a few minutes ago on. You're trying to understand what's happening with an ISP or something like that. So I'm here in San Francisco at our office and I'm connected to our lab network here. One of the first things you can do in command is similar to Slack or Teams. You can just hit at to get started.
Speaker 3:Slack or Teams, you can just hit at to get started and when you do, you can like scope it down to a particular piece of hardware, particular network, a particular client. But why don't we select the network here? That's our primary test network and you know, maybe you want to say something like when was the last time ISP quality was below 95% or something, and you can just like type it out as I'm just exactly as I'm speaking, and the cool thing about command what it's going to go do, that will seem very simple is it just went and combed through all of the logs on this ISP and actually just gave an answer back and here the answer is just saying hey, this was this particular time it was below 95%.
Speaker 1:And it has to know what 95% is also, so it has to understand and be able to interpret what you mean, how you mean it, and then apply that to the network. And I mean, it took what? Three seconds maybe, to respond. No, no, no, it did not take three seconds.
Speaker 3:It took three seconds to render. No, it didn't this one. It took about 560 milliseconds.
Speaker 1:If it took three seconds, I wouldn't even be here.
Speaker 3:I would never show it to the world.
Speaker 1:That's so great, that is so great man 560 milliseconds.
Speaker 3:But after that you were kind of saying something about you know switches early, right, You're kind of looking at particular switch. So let's say, today your job was to add a bunch of things to switch ports, because maybe you're adding security cameras or maybe you're adding access control or something like that and there's a particular switch you care about. Right, you can just go to that switch. I'll pick a random one, you tap on switch and then you which you care about.
Speaker 1:Right, you can just go to that switch. I'll pick a random one. So you tap on switch and then you choose. Yeah, I just hit add and I just selected.
Speaker 3:I'll just select to that switch one of our test racks and I can actually just say, like show me a port diagram. And again what will happen that will seem innocuous is it went and actually just wrote software entirely on the fly for a port diagram for that switch in line.
Speaker 1:It didn't have that the picture that we're staring at wasn't. That's not some pre-made photo, that's not.
Speaker 3:Oh this is real life software that you can switch, just like you can. I just switched it to another switch. Right now it's fully usable.
Speaker 2:This is a unique piece of code that's only running right now inside this command. That's right, and yet it's still. It's workable. You drop down to a different switch. It's a different switch. It does look like your other dashboards.
Speaker 3:I'm guessing.
Speaker 2:Since you already had the code, you could reuse some of that code.
Speaker 3:Yeah, so what's happening underneath actually is really interesting. One part of command is a model that understands our hardware, our architecture and kind of our entire system our architecture and kind of our entire system. Another model that actually understands how we write software and how we design software. And, keith, that's where our design system is, what you're seeing here, because we have a design system that we try to maintain consistency. Where a table is a table, a port is designed a certain way, we use certain colors, and so it's trained on our design system as well, so it knows that it should look like meter software. That's awesome.
Speaker 2:But it's not just meter software because as you switched, it's getting live data from each of those ports on the new switch.
Speaker 3:Yes, that's exactly right, and that's the other really amazing thing that our team ended up building. As far as the architecture is concerned, the software that's written is like an engineer would write. That's fully configurable and you can kind of have dropdowns and filter and search and all these different things, but it's also pulling real time data from every piece of hardware, all at the same time, all the same time all the same.
Speaker 3:So maybe this is what you care about, because you're adding a bunch of um, you know, security cameras or something um. If you remember earlier I was mentioning one of the awesome things about working in the networking lab is people peering over your shoulder, whether you want.
Speaker 1:That is such a just such a kick-ass way to say that, because now I I mean, I know where you're going with this, right, but it's like that's all. I mean this, keep going, that's awesome to have that visual of someone peering over your shoulder and realizing that you do not need that anymore because you're that's incredible breathing down your neck you just feel it right, as you said, that yeah, so so maybe you know you just want to care about, just feel it right, as you said, that, yeah.
Speaker 3:So maybe you just want to care about this one switch right now. You can just actually pick it up and drop it here in the green zone here, and this is a full multiplayer area and collaborative.
Speaker 2:So Anil just moved the port diagram that he made in the previous step over to the right side of his screen. A little worksheet, a little empty area and it's now sitting over there in a shared area and it's just going to live there.
Speaker 3:Yeah, it's going to live there, no-transcript, and you're at work and do your work is some SEV1 issue comes along. As you know, at Meter, if any of the Sarahs say anything, it's SEV1 all the time. That's all they hold the power they do. So maybe we'll go like a particular client, as we go to like actually like Sarah's MacBook Pro, yep, and she's saying something's going on, things are not working. So you can literally just act and get to Sarah's MacBook Pro and say tell me about the help of this client and give me a five-bullet summary. What's going to happen is two things. One is it's going to comb through all of the data about Sarah's MacBook Pro and give a summary again at the speed of a Google search, not three seconds.
Speaker 1:Yeah, it's not three seconds y'all. That was a terrible miscalculation on my behalf.
Speaker 3:And it just gave a summary about hey, for this MacBook Pro, this is the Mac address and this is what I'm seeing across signal strength, noise level, usage data rates and SNR. So you start and then it wrote software that gives the entire help and, similarly, you can pick it up and drop it here. Because you can pick it up and drop it here? Because you know Sarah's MacBook Pro is the most important MacBook Pro at meter, absolutely. You're always paying attention to it. So that's the other really interesting part is making debugging easier, getting information really fast, but really what? The underlying truth of this is that it saves time, right? Normally, if I'm going to go look at all these logs, it's going to take some time for me to understand what's actually happening.
Speaker 1:Well, the other thing that I've been a little bit vocal about is that there's all these ideas of what AI can look like and chatbots and all these things of ways to interact with the system, and my biggest bone to pick has always been but you have to learn how to talk to it. You have to learn how to ask it a very specific question. You have to understand the syntax that it uses, because you can't just have a conversation with it. What you're showing is that you can just have a conversation with your network in whatever way that makes the most sense to you. It can be as technical as you want or it can be absolutely as non-technical as you want, and it's going to give you information that I wasn't ready for it to show SNR. I thought maybe it showed data rate or you know, last time it went down, I wasn't ready for something that was actually important to be shown, because it's a lot of times people don't think about that.
Speaker 1:It's you're showing off a way that allows us to truly interact with a network on our terms, our terms, and this is this is incredible to me um, and it's there's so many shock waves of this that I'm looking at and I'm thinking about that as this ripples further out. These are the things that people in, at least in my mind, these are the things that I thought. This is what that ai revolution in networking is supposed to look like. You're showing it to me, and it's the first time second time that I've seen this because you showed it to me a few months ago. That's the first time that I'm actually seeing this, and this isn't some canned demo. This isn't some. This is real life. Asking functionality of a network and having it display information, information to you. What were you gonna say, keith?
Speaker 2:I. I think the beauty of this it is that Anil's and his team have trained their model on their stuff. It doesn't have to be a chat GPT that takes anything from the internet. It's you know your own software, you know your own firmware, you know your own coding, you know the data that's included and where the where all the packets are going on the network. You don't have to train it to answer. You know what's the height of Kilimanjaro. You can let it focus and I think that might be a part of the speed process. But I think it's answering what Drew was looking for. It's valuable because it's answering my question about the things I care about.
Speaker 2:Yeah, contextual, I think it helps that it was trained specific to do this one thing and not just to do anything.
Speaker 1:But it's not so and, anil, maybe you can answer that question from a quote-unquote training perspective. It's not. I don't feel like this is necessarily trained. I think this is a constantly evolving thing as your network grows. Is that fair?
Speaker 3:So it's a combination of both right, it's built and trained and the models on our architecture, how we write software, how we design software, our backend, those things but at the same time, it goes and gets real-time data about all the clients and all the piece of hardware, because what you want is the real time data. And the architecture where we ended up with command is super interesting because we didn't take any customer data and train it on that, because we really care about privacy. We think it's incredibly important, especially in enterprises. It's only trained on, as Keith was saying, our architecture, our software, our design and what it does is goes and grabs real-time data about each of these clients and each of these devices on the network. So both of those is the combination where we ended up.
Speaker 2:It also means it has to have some other extra intelligence. I don't know where it's coming from, but it had to assume. For signal strength you give a range, which means that was the range over some period of time. So you had to put a time variable into the equation without me specifically saying over the last three hours.
Speaker 1:Neil is smiling and grinning from ear to ear, showing off his I just like he needs to ask the question.
Speaker 3:It's Keith, because he can see everything right. He's been around for so long. So, yeah, keith, it knows so many things about time. It knows things To give you a little bit more under the curtain too. It has its own calculator too. That we built under the curtain too is like it. Um, it has its own calculator too, that we built a bunch of things that are happening in the background so that it can actually do things, and so, from time to calculators, to understanding even how to write, write markdown there's a lot of things that are going on uh, in here, okay, so so then step two, right, yeah, you've shown us information about the network.
Speaker 1:That's awesome. What can I do with that information? Can I configure devices? Can I?
Speaker 3:Exactly, exactly. So, okay, you solved Sarah's problem, because, again, that's the number one MacBook. Now you're getting back to your work and maybe to your earlier example, drew. The next thing you're doing after you these security cameras and these action control systems, is you have to go configure VLANs, and there's a prompt that I'll use that's already been typed because it's been part of the testing since we started, so now it's like a funny prompt that I use at every demo. This is for Peter, this is for Peter, your chef.
Speaker 1:I love it.
Speaker 3:I wish it was for Peter. He was one of our researchers that I think was hungry at that time because he'd been working all day and all night. So maybe you're saying something like create a VLAN, call it Lobster, say it's for cooking, enable it and give it a certain VLAN ID, and when you hit enter, there's some interesting things that are happening here. Command will go understand that query that we're saying, and actually it just wrote out an entire form that's been filled out. So what you're seeing here is a form that normally would be in one of our dashboards where it'll just fill out entirely and you can go look at it and edit it. You know, you can kind of say, like it's not for lobster or whatever the case might be. You're kind of editing it and then you can confirm by saying I actually want to submit this.
Speaker 1:There you go, okay.
Speaker 3:And then there's another step, because we believe mutations are expensive and we want to make sure the users understand what they're doing, just like you would anywhere else. You want to have a Of course You're going to see how it's going to impact the network and then, when you execute it, it will actually go create that VLAN, and it was that fast across the entire platform. You just said what you want. So getting information and then taking action as well, and again, those are the two great powers of a command line.
Speaker 1:So to go off script, then would it be possible for you to type in create an SSID called Lobster and tie it to the VLAN Lobster? Is that something that you can do and press?
Speaker 3:it. Yeah, you can. I just want to make sure I'm not bringing down the network. But yes, you can.
Speaker 1:Well, yeah, I don't want to ask because I know you're actually on your network, but if you get a thumbs up back there, I'd love to see that.
Speaker 3:Yeah, it can do SSIDs, firewall rules, dns security, port forwarding, rate limiting. Just, it understands our entire architecture and I think it's an important point to reiterate here is you, uh, open with asking us about our architecture and our stack?
Speaker 1:absolutely, that's what command understands every okay, okay, I mean every bit of your architecture, or like 90 of it or 60. I mean, what are we talking? I mean, is this every bit of it? And Is this the real deal? That's what I'm trying to figure out.
Speaker 3:Yeah, yeah, every bit of it. And there's two times it continuously updates itself. One is when we change that architecture, there's a CICD pipeline that gets kicked off. Okay. Same is when we change our design, then our CIC CD pipeline also gets kicked off, so the command can actually learn those new things. So maybe I'll say like there's usually some gap there somewhere between 24 to 48 hours so the new model comes out, so that understands as we make changes. So let's say, you know, recently we announced a bunch of stuff around multi-WAN and high availability and all these things. As that feature goes in and it gets tested and pushed out, there's a gap of 24 to 48 hours before command knows all of it, because the models have to get trained again. But that's a CICD pipeline that gets kicked off.
Speaker 1:We've been waiting 10 years to see this, anil, and you're telling me we got to wait an extra 24 hours.
Speaker 2:I don't know if that's going to be acceptable or not. I have a question for you, neil Sure. Could you ask command the same kind of thing but say I would like to do these things, but tell me if it's going to break anything. Ooh, nice. So yes, you had an execute button there. By the way I wanted to say make it so, but execute works. But can I have command help me not make boo-boos.
Speaker 1:And if you're listening to this Anil's typing I would like to add 10 VLANs. Is that going to break anything? And that's literally what you wrote and it responded. Adding 10 VLANs to the network primary should not break anything. The current network configuration supports additional VLANs. There are no existing limitations or constraints that would prevent you from adding more. The network's current load and performance metrics indicate that it is operating well within capacity data. Eat your heart out, man. This is like wow, what? Okay, so this is the problem that I had before.
Speaker 3:If you can do anything like we're, you know, I do want to I do want to answer that a little bit. So, um, one of the great inventions that happened in the last 15, 20 years that's not that noticed is actually when you used to go to google and type something in. It would say did you mean, yeah, and there's this great engineer at Google that actually just came back after leaving for a few years. His name is Noam Shazir and Shazir and his team built, did you mean? And it's one of the really hard problems to solve, because not only are you saying here's what I can understand, what you asked, but did you actually mean this? And so that's actually what's coming with command.
Speaker 3:Next is taking some of that work and how do we apply that to networking? So, totally, here you drew that. The blank canvas is scary and not that inviting. So we're actually going to do two things. One is, when you open with command soon is that we'll actually give you queries that you should be asking, and then, when you do ask a query, we're actually going to try to suggest a better query to get to the thing we think you're trying to get to.
Speaker 3:That's awesome so we are building on that. We totally hear you. We don't think what you're saying is like to that's awesome. So we are building on that. We totally hear you. Uh, we don't think what you're saying is, um, like anything out of the ordinary. We think it's an important thing. Uh, we want to make it really great before releasing it, but we'll fast follow that into command so okay, so so where does this leave the elephant?
Speaker 1:the elephant in the room question where does this leave the current status of network engineers who have shelves full of books and walls full of certifications that they've spent so much time learning, I mean, and I think about programming, I think about you know, I think about lawyers and tax code and everything else, and we're at this turning point where anything that involves a whole lot of existing knowledge based on libraries of books and understanding is now quickly catching up to a point where I don't want to say you don't need it anymore, but you're not going to need it soon anymore. But you're not going to need it soon. You can. It's great to have, say, the people that are fighting back against it. But you're entering this time where this is like the Google scenario. It's like, well, you need to learn all of the knowledge and the younger generation says, no, you can just Google it. And you say, yeah, but what if you need it when you don't have access to Google? And they say you will always have access to Google.
Speaker 1:But you know, we're at this point where where the network engineering teams have looked at, have thought about things like this and they've thought how is this going to change our jobs? What? What's the answer there? I mean cause, this is the. That's the tricky question. It's not. It's not replacing people's jobs, that's not what I'm getting at, but it's replacing the need for knowledge. What effect does that have on certifications and understanding and things like that? And, keith, I want that question to go to you too. What do you think about that?
Speaker 2:I mean, I've been in the certification game, I collect them. I have like 115 now. It's just a hobby, but I think it means we need more education because you need to understand the concepts. You might not have to have the skills. I don't need the muscle memory on the command line, but I do need to understand how BGP runs on a router. I can't. The base knowledge hasn't gone away.
Speaker 1:Yeah.
Speaker 2:I think that the keystroking I don't have to know, that I don't have to remember things. I can look those up. It reminds me of my teacher in seventh grade said what do you mean? Use a calculator? Are you going to have a calculator in your pocket your whole life? Yes, I do. So. There are that whole. You can Google it for information. I don't think you can Google knowledge.
Speaker 1:I think that's the difference.
Speaker 2:I have the knowledge but, we just have some tools to help us go a little better I, I 100 agree with keith.
Speaker 3:Uh, our goal with command is to make it easy to get information and faster to take action. I think of it a lot like um. You know you brought up certifications, both of you, I. I think of it a lot like you brought up certifications, both of you. I think of it like an open book test, whereas when you're in school it's closed, but in real life it's open book, and we're hoping to open all the books for our customers and network engineers. But you still need to know what to do. And will we make it easier? Will we help you get there faster, and should it be exactly what you want? Absolutely, but real-world networks are open book tests and hopefully, with Meter, your books are even better.
Speaker 1:Yeah, and I think this goes back to I love it as soon as you said when you talked about either the proctor, the lab assistant or the person standing over your shoulder watching you do something. You have that. Now you have some. You have a tool that gives you the ability to not just be that open book but to deliver the information to you when you need it the most. It's just, it seems like this is going to from a revolutionary perspective in where we are with networking. This is such a great way that people who have the understanding but maybe don't have the memory retention of all the key codes and the line items in CLI, this is going to enable so many people to do so much more. Um, and then with the you know, with the bumpers or the guidelines that are set up to make sure that, well, if you do this, it's gonna break this and this and this. It's such, a, such a neat way. So so you've showed us that you know you guys have access to the software, or you know, to everything up and down the meter stack. You show us that we can receive information using command. You show us that we can configure the network and to what, to what that, that extent is and those limitations are.
Speaker 1:I'm not even sure yet, because it's this is. It's continuing to evolve. What else about command? You know in the in the next couple of minutes, what else about command are you looking at? Because you mentioned something about building your own software and I know that was key with this. This it's. This isn't just a tool to tell your network what to do and receive information. This is a tool that's much bigger than that. Can you talk about that a little bit?
Speaker 3:Yeah, absolutely so. There's three important parts of command, like you mentioned. One is getting information really fast and that can be across any part of the stack ISP, security, firewall rules, whatever. You're just trying to get information. Second, is you want to take action. That might be something complicated saying create this as this ID, make sure it has an auto-rotating password and rotates every day at 2 am and make sure it's WPA2 PSK.
Speaker 1:Prioritize traffic, for whatever it is, you can say that it takes action. Prioritize traffic for whatever it is, you can say that it takes action, the last part of what's command.
Speaker 3:what's important for us is tying back to what I mentioned earlier on the evolution from CLI to dashboards to where we think is the future is having software written for you entirely on the fly. So we saw some components right. So we saw it build switchboard diagrams for you on the fly health software forms.
Speaker 3:But to show the power of it, I want to go all the way through so maybe you say something like I am a network engineer, give me a dashboard I can share with my CTO, and what it's actually going to go do in earnest is write out an entire dashboard just the way you want it.
Speaker 1:Just the way your CTO wants it.
Speaker 3:Yeah. So topologies, uptime, network throughput devices that are connected, youlans for security, error rates report and channel utilization all fully written at the speed of a Google search. Wow, and those are clickable, clickable, searchable, everything, everything here. I'm searching this thing right here. We can search for appliance, whatever. I can then drag these over entirely and start building out how I want it to be.
Speaker 3:And this software that we have here is when you're in canvas mode. That's where everybody can jump in and you're all collaborating and maybe Keith will drag in wireless and you know you're going to drag in some network throughput stuff. And then, when you go into dashboard mode, we can actually just start configuring what we want the dashboard to look like. And it's actually just going to be full dashboards that you essentially have a software engineer sitting next to you and you can just share this out. You'll get a link. It's all behind SSO and role-based access control and in our case it'll say hey, only people in meter will be able to view this canvas. You can just copy that link. But you can create any number of them. You know you can say Drew's dashboard and create that and share it out. So get information, take action and have entire software written for you on the fly, just the way you want it every time.
Speaker 1:I'm sorry. I didn't think I was going to be shocked again.
Speaker 2:I didn't think I was going to be true, but I just seen you do it.
Speaker 1:This is nuts, man. Where do you go from here, anil? I mean, it's just like it's. This is great, I mean, and again I'm trying to, I'm trying to go back.
Speaker 2:You made Drew stop talking. That's really hard to do.
Speaker 1:I just you know he's at a loss for words you start with the ISP. You get the ability to pick and choose the ISP that you want to be a part of. Then you get to you know, select the network components, or the network components are selected based on your needs and what you do and how you're doing it. It ships, it's installed, everything's working, and then you've got this ability to interact with your network unlike anything we have ever seen before, and it's taken the complexity out of it and given it to you as simple as you need it, but still has all the bit of complexity underneath it, should you want to go in and make changes. I mean, and the reporting side, and I really I I don't know where you go from here, and I'm sure you do. I mean that's in another two months. You, you're going to have something to show off, I'm sure.
Speaker 1:But this is, you know, this is really, this is dude. This is really great. I mean, from friend to friend, this is so badass. Like I am, I am so happy for you, but I'm so happy for our industry that something incredibly innovative like this is is becoming available. Man, and if you look at, if you look at what's happening in the market. You look at what's happening in the industry and you look at everyone trying to figure out how to. What are we going to do? Where are we going to go? How are we going to shift the money? How are we going to shift the focus? How is it going to go to ai? What does that even mean? You are showing us something right now that I have that I. I can't imagine the look some people's faces when they start to see and understand what this is. So kudos to you and the entire team at Meter. This is insane. I love it I love it.
Speaker 2:I think what you're trying to say, Drew, is you didn't know you wanted it until you saw it, but now you have to have it.
Speaker 1:You have to have it. I mean, this is like it's incredible, man. I think about the simple functions of configuring a network. I think about the advanced functions of configuring a network. I think about everything that I've had to learn to do what it is that I'm trying to do and being able to not automate isn't the right word but I'm able to use a tool that just makes it so much quicker for me that it's going to allow me to focus on other things that are important, and I think about CTOs and CIOs and the staff that they have, being able to interact with this and truly start to focus on information technology, not on configuring devices and supporting devices and administering devices. I mean, this is an incredible shift and I'm so happy that you let us talk to you about it. This is really cool.
Speaker 3:Yeah, I mean I just think you know the infrastructure decisions we made, the alternative choices that we took and the paths and the pain to build good products led us down here, and we really want networking to have the best tools in the world and not look over and look at other industries and be like, why don't we have something cool?
Speaker 3:We should have the things that are at the forefront and, at the end of the day, we hope our customers and our partners can see the value of meter even more and, like you said, I do know where we're going and we will come back and show you more. But this is just the start of what we're doing. But it's a culmination of many years of work and what we think we can do in a different, special way that currently I don't think the industry is set up to do, for a myriad of reasons. Currently I don't think the industry is set up to do for a myriad of reasons. But these are the things we want as well and ultimately, faster, secure, better networks for our customers, but also, at the end of the day, really great products. We just think networking should have the best products in the world.
Speaker 1:What you've just seen is an incredible sneak peek of what's coming up with Meter as we come up to the top of the hour. I appreciate everybody who's listening and tuning in such fantastic guests. Keith, it's always good to see you, and Neil it's. This is just so innovative and fun. Can't wait to see what comes next. If you need more information, visit metercom. Look for one of the Sarah's on LinkedIn. They'll be happy to help you. Look for Mr.
