ERP068 - Evolution of windows virtual desktop

We built a new product and launched it on March 10th of 2020. And that product was called Nerdy Manager for WVD and really was designed for very large scale, you know, up to tens of thousands or even hundreds of thousands of users. And March 10th, I'm sure as many of us will remember, was just a couple days before things went rapidly bad in terms of the pandemic and all the lockdowns. And and as you said, we saw just unending demand for organizations wanting to both use our product and have us help them deploy it and and manage it. Welcome to Evolved Radio, where we explore the evolution of business and technology. I'm your host, Todd Kane. Today on Evolved Radio, I'm chatting with Vadim Vladimirsky, CEO of Nerdeo. Vadim and I discuss our shared experience in the early days of remote delivered desktop computing. Vadim was an early visionary on what the technology could eventually become. He built an MSP and set to work in building a solid desktop solution for his clients. His success with the platform led him to spin out Nerdeo from his MSP and has built one of the channel's leading solutions for serving and managing Windows virtual desktops. We talk about the early issues of the technology and how things have matured. Also, as the name of the company implies, this episode gets pretty nerdy. So get ready to go deep and better understand the power of Windows virtual desktops past and present. If you enjoy the show, please consider leaving a rating and review in your favorite podcast app. It really helps to spread awareness and bring more listeners to the show, so we can share the message with more of the community. Now, on with the show. And joining me on the podcast today is Vadim Vladimirsky, CEO of Nerdeo. Welcome, Vadim. Good to good to be here. Thanks for having me. It's great to have you as well. I'm sure some people in the MSP space will certainly recognize your name. It's tossed around a fair bit and definitely attracting some industry recognition with your solutions. And interested to kind of dig in and chat a bit more about this and sort of how you guys got into the space that you're in now and the solution that you're offering for for the channel as well. To kick things off, if you could, maybe just give us a bit of your background, a bit of your origin story and and how we arrived at Nerdeo. Sure, so I'd be happy to tell you a bit about my background. So I I started in IT and technology, I guess when I was back in high school when I was going around people's homes and helping them get their Wi-Fi network set up and and printers going and things like that. So that's sort of how I fell in love with IT and and doing things with technology. I studied computer engineering in college and business as well and was always fascinated by the business of technology. So I in my career, I wanted to really combine the two. So I spent several years after college working for an MSP. And then a few years into that experience, I actually started an MSP with the bosses that I was working for at the time. So the two founders of that MSP, we started a new company, this was in the mid 2000s and my vision back then was to have an MSP that will have a service offering that will provide hosted virtual desktops to small business customers all over the country. Right, which was a a pretty unique concept back then. There was no cloud. Amazon S3 came out only the year after that. But that was always the vision. And and the reason I was really so excited by that vision is because I I was exposed to Citrix and Windows Terminal server back when it was NT version 4.0 back in the late 90s. And and really thought the technology had a lot of potential, especially in a small business setting where you don't have an unstaffed IT person. If you can have someone who can host and manage all of that for you. It really could provide a lot of business value. So, that's how I started in IT. I was, you know, started an MSP in 2005. In about 2016, 10 years into that, we we created a a fairly large MSP with, you know, many hundreds of customers all over the world. And what we realized is that the technology that we created to operate the business, you know, the provisioning, the management, the billing, everything that that we created to to use internally was really something that other MSPs could benefit from. So we decided, hey, let's create a platform that other MSPs could use to build a similar cloud practice. So cloud was obviously getting very popular. At that point and we decided to provide sort of an an easy button and a turn key solution for MSPs who wanted to grow that type of a practice. That's how Nerdeo was born. It was actually born inside of the MSP that I was running. And eventually was spun out of that MSP and we exited the actual MSP business a few years later. You know, that I guess that's how it all started. And and has evolved significantly, we've added some products and things like that. But but that's really the the quick background. In in two minutes or three minutes. Okay, appreciate it. Yeah, it's it's an interesting sort of parallel with so many other stories of services and products that get created in the MSP channel is often it's developed on scratching your own itch as it were, right? It was something that you developed for yourselves. And you're like, hey, you know, other people could use this as well. It's such a common story in our channel as well. Absolutely. You mentioned that originally, like you had this vision. Sort of way back when. That's that's pretty early to imagine sort of a cloud delivered desktop in the 2000s. I come from a similar background, I'm a former Citrix administrator around the same time. You know, the the NT systems were sort of the first systems that I was building Citrix on top of. So I certainly saw the power of it, but didn't necessarily envision this as a sort of a complete solution. Like what was it that sort of led you to believe that this is something that that could be delivered in its current form back then? It's a great question. I I was actually told I was crazy by several industry experts back then. I can imagine. Yeah. Yeah. Yeah. So the way this all was born is as I think we had a customer that was looking to do some sort of DR. They wanted a way of being able, you know, to have a DR solution. They were they were a fairly small company, they were logistics kind of brokerage customer. And, you know, we were trying to think of like how would we architect this for them. And we thought, well, what if we get a few, you know, pizza box servers, rack mount servers and we put Windows 2003 with terminal server on top of it. And and set it up in a data center and and get them to to run on that. So we piloted it with them and and it was, you know, pretty successful. And that was I think the original idea of, well, you know, if if they're using it for backup and DR, you know, why not use that in production? They have a decent, you know, that decent connectivity. So performance was good. So, so the idea was born out of that. Obviously, crunching the numbers, it was hard to justify a price point with all the infrastructure that was necessary plus the data center costs. To actually put people on physical desktops, you know, on physical gear with terminal server, Windows 2003 terminal server at the time. So I started looking into virtualization and, you know, VMware infrastructure, I think it was version 2.5 back then. Was all the rage, right? This is before server virtualization was was widely accepted. And I said, well, you know, why don't we just use that, you know, we'll we'll have a few physical hosts and then we'll have a bunch of terminal server VMs on top of it and and deliver it like that. And ran the idea by a few people whose opinion I respected in both the terminal server world and and the virtualization world. And they said like there's no way the density will never be high enough for you to ever make any money. But I was young and stubborn and I said, no, this is going to work. I really pushed through it and and it worked. It was really, I mean, it worked well, we were profitable almost right away. You know, as soon as we signed up a handful of customers, the infrastructure was paying for itself. The hosting was being paid for. And there was good margin to be made. So, you know, by by 2006, 2007, we saw that there is really a model here that can be used to scale a pretty large business. There wasn't anyone else in the market doing it, right? There was obviously lots of people doing virtualization for internal use or for enterprises, but nobody was offering a hosted virtual desktops, certainly not in the MSP space, but even not sort of on the market. We we didn't see anyone else doing it. So we had to really start. Yeah, it's it's interesting, so that that that first iteration was multi-tenant to begin with then. It was. It actually started as multi-tenant and then around 2009 was pretty pivotal point in in the evolution of our technology stack. So up to then we were using VMware for server virtualization, so getting the VMs that are running terminal server to run. And then the presentation layer has always been RDP and it was always terminal server in a multi multi-session type of an approach. And users had to have a local machine and their virtual desktop session. We used to set up the background as black on the terminal server and keep it as blue on local, so when they would call with a support ticket, we would say you're on the black screen or on the blue screen. Just so we know what context they're talking about. And the reason they needed a local machine is because anything graphically intensive, any anything with audio, any video, they just needed to know, okay, you minimize this and then you do things locally. What happened in 2009 is VMware, which was in version four of VMware view at the time, licensed a display protocol from Teradici called PCOIP. And they integrated it into VMware view 4.0. And you know, I started I was reading about and I thought, oh, that's that's interesting, there's this concept of a zero client and and video is supposed to work well. So we tested this out towards the end of 2009. And it really did work pretty well. I mean, it it ate a bit more bandwidth, but the quality of the connection and the ability to scroll through PDFs and, you know, listen to videos and watch things on the screen was just night and day between what was available with RDP at that time. So we went from a primarily multi-session delivery to a single session VDI delivery with VMware and PCOIP. And that was kind of an inflection point in our business because now customers did not need to have a local device at all. And primarily they were buying these inexpensive zero clients that had no operating system, you know, no spinning parts, it's not even a thin client, there was not even a Linux or embedded OS, it was firmware only. It was usually integrated into their display. So it was like a 24 or 19 inch monitor. And that's all they needed to get to their Windows desktop with very good experience. And that's how we went to single tenant at that point. Or single single user. It's interesting because like I have experience in this as well, like being a former Citrix administrator. We started to sort of be pushed by a lot of the vendors to to move towards this VDI model. And and I remember being in lots of these sales sessions with clients and saying, you know, this the VDI is an amazing solution, here's all the cool things that it allows for you, work from anywhere, all of your information is completely portable, all of this. But the hinge point in a lot in most of those cases was just the sheer cost of infrastructure uplift in order to give you sort of the the raw infrastructure necessary to get sort of a critical mass of people working on there. And it was extremely sort of memory and CPU, especially memory, but CPU intensive, you had to get these blade chassis with a ton of equipment. You're like, okay, so spend $750,000 on infrastructure and then we'll be able to get you on to VDI and everyone's like, oh, god, I don't know, this is really expensive. So I think I feel like part of the success of your model was being multi-tenant first because all of the conversations that I was having in the in these instances were infrastructure for a single sort of mid-tier enterprise organization. And there was a lot of companies that did buy into it, but it was really, really big companies that had the infrastructure spend to be able to get that infrastructure there. Do you feel that that was sort of a necessary component of the success of being able to distribute the cost across multiple organizations and not have to have a single organization absorb that the the infrastructure cost? Absolutely. I mean, without it, as you mentioned. Small businesses would not be able to afford this type of a solution. One decision that we made really early on that I think was partially responsible for for how well this was adopted. was we decided to be multi-tenant across the physical gear, so compute, storage and networking up to layer two. So we basically did layer two VLAN segmentation for all of our customers. So each customer, even though they were sharing the compute infrastructure and the storage infrastructure, had their own VLAN, which meant they they had their own active directory, they had their own file servers, their own terminal servers, their own VMs. None of that, everything above the network layer was really segregated. And what that allowed us to do is really be flexible with what each customer needed. Unlike many of our competitors that we saw come on the scene later, they were trying to solve the problem of cost by not only being multi-tenant across the hardware, but also trying to leverage the same domain controllers, the same database servers. They were trying to be multi-tenant all the way up the stack. And in those scenarios, number one, lots more outages because you have one customer can impact others. You have noisy neighbor problems. You have, you know, you make one change and it has global impact, so it's difficult to keep things updated. So there's just lots of operational challenges that we luckily never had to deal with. Because we designed the system in such a way that if a customer needed something custom done in their active directory schema, we could safely take a backup, make the change and not think about what impact is that going to have on anyone else. Because nobody was ever using that same active directory as as this particular customer. But getting back to your question, yeah, the the way to, you know, multiply and and distribute the cost of the hardware across customers is is absolutely key to making this kind of model work. You know, we went through lots of iterations. We actually found that compute and memory were not the biggest cost components, right? They were significant. But they weren't the biggest. Storage was by far the biggest thing that we had to deal with because in order to take advantage of all of the fancy server virtualization things like the V motion and and DRS and all of the high availability stuff. You need a shared storage system, which means a SAM, right? And Sans, as as everyone knows, are super, super expensive. So we started out with EMC. That did not work out well. We went to NetApp, that worked pretty well for for a number of years. And then eventually we settled on Nimble storage that was acquired by HPE. And that really allowed us with the all flash arrays and the way we've segmented things to really build up a very sizable number of desktops with a very small amount of equipment. So, you know, our data center footprint kept shrinking, even though the number of customers we were serving kept growing. This is interesting, I'm getting all kinds of flashbacks. This journey mimics a lot of what I experienced along the way as well. Even the vendor names in sequence are the same, so really, really interesting. I want to talk a bit about like a lot of people are very familiar with terminal services and remote desktops, sort of the the multi-tenant and just sort of screen cap push to to a local desktop. Not as many people use a virtual desktop. And I wonder, you know, if you could maybe speak to what you see as the advantages of a true virtual desktop versus a terminal service solution from a either from a provider standpoint and certainly from a user standpoint. Yeah, and and if I just, you know, may sort of align with you on the terminology. So I think when you talk about terminal server, you know, we typically call it multi-session, right, when you have one OS supporting multiple concurrent users and virtual desktop just a single session OS that one user is connecting to. Yeah, like this is my version of Windows 10 every time I log into it, I get the same persistent session versus I've carved out a piece of the OS and I'll give you this piece, but you're running in parallel with everyone else, right? Perfect. Yep. So we're aligned on that. Because we live in the WVD world for the last couple of years of Windows virtual desktop. They call that pooled versus personal. So if I use that terminology, you'll you'll know what I'm talking about. So, the use cases for the two scenarios, pool versus personal, are fairly distinct. For example, typically a pooled desktop is appropriate when the user needs dedicated performance, so you want to basically be able to guarantee a certain amount of CPU, RAM, disk and IO associated with that on a per user basis. So think about, you know, like an executive or somebody who doesn't want to have variable performance depending on who else happens to be on the system. I would say that's scenario number one. Scenario number two, it's it's about sort of software or security isolation. Because a user is running by themselves in a single VM, you can guarantee that all of the software that's installed and everything that's happening is is really being utilized by that user and that user only. Whereas in a pooled scenario, you may have the same experience, multiple users are logging in, they have their own virtual session, but they're sharing the resources, which means one user in theory could impact all others. And also, any software that's installed on that machine, you have to go through hiding that software to make it unavailable to others. As opposed to in a personal desktop scenario, it's not installed, it's not there, and if it is installed, it's there and it's only impacts one user. The reason, you know, customers and and and MSPs choose to go with one version or one model versus the other. And by the way, in WVD, you can mix and match the two in the same deployment, there's really no no reason not to do that. You know, with pooled scenarios, primarily the cost is a fraction of a personal scenario. For obvious reasons, because with a personal desktop scenario, you're allocating a dedicated VM, let's call it, whatever, $100 a month, right? So that VM cost you $100 a month. And that user is going to consume $100 a month whether or not they're using that VM because it's been assigned to them. In a pooled scenario, not only are you getting multiple users on a single VM, so the density is more than one on one to one, which is a savings right there. But also, if not all the users are logging in at the same time, you can save because the concurrency is generally lower than 100%. Not everyone is always working. Which means you can get away with less capacity per user because not all the users are using it all the time. So it's it's always almost almost always a cost driven decision. And the cost differential is significant. I mean, it could be a difference between, you know, $50 per user in a personal desktop scenario with less than $10 in a pool scenario. For what appears the same to the end user, same application, same experience, it just it's not my dedicated personal machine. It's a pulled machine with everyone else. So would you say based on sort of the way that the Windows virtual desktop works, we're sort of away from that that history that I experienced in in sort of the older versions of this of the virtual desktop being a sort of assigned and provisioned for a particular user in order to have some level of persistence rather than sort of a standardized desktop. It's sort of less necessary because of the sort of the the maturity and the features that are now available in that pool scenario. Yes, yes, there definitely have been features that were added to the pool scenario to emulate that virtual desktop single user scenario. One other thing I guess that I didn't mention is previous to Windows virtual desktop, WVD. Whenever you're talking about pool scenario, you were talking like you said terminal services or RDS. Which meant a server operating system. Which is a key distinction I no longer think about because in WVD, it's no longer server server OS. But but that was significant. Because if you are doing a one to one, a personal desktop in the, you know, old days, you're giving user a Windows 10 or Windows 7 or Windows 8, a desktop class operating system. Whereas if you're doing multi-session, by necessity, you have to use RDS, which is a role of Windows server. So software compatibility could be could be an issue, drivers, all of those things came into into the decision matrix. That has been eliminated. Why? Because Microsoft created a new version of Windows 10 called Windows 10 multi-session or EVD, stands for Enterprise virtual desktop. And that version is identical to version, you know, Windows 10 Enterprise, which is the single user version, but it just allows multiple users. So there is no longer the consideration of do I go with a server OS that looks like a desktop or do I go with a true desktop OS? In WVD, you don't have to make that choice. And that's when the the the cost decision comes in. Do I give someone dedicated resources and pay more or do I pull the resources and pay less? Now, how, I think going back to your question is how did Microsoft make it so in a pool scenario, things look the same to the end user as they do in a single user scenario. And that primarily has to do with profile containerization with a technology that they bought back in 2008 as they were working on on WVD called FS Logics. And, you know, if you think about a pool scenario, what that means is a user is logging in and they're logging into any one of the available machines in the host pool. Which means that there's no guarantee that they will be on machine A today and and not machine B tomorrow. And what that means is every time they log in, their profile loads, and if you are on machine A yesterday and you save the file to your desktop. When you log into machine B today, that's a different machine, different C users, username desktop folder. And your file won't be there anymore, which obviously makes virtual desktops unusable to typical users. Who do have customizations, who do save data, you know, that was a a challenge. Which was solved by various technologies called user profile disks, roaming profiles. Various technologies that existed before Windows virtual desktop. The drawback with a lot of those technologies was that they had limitations, they didn't support things like Outlook caching, indexed search, OneDrive. All of those things that users really like were not available when using it in a in a multi-session kind of way until FS Logics came around. FS Logics basically takes the user's profile and containerize it into a VHD virtual disk file that sits somewhere on the file share, whether it's an Azure or on the on the on the VM. And it gets mounted under the C user's folder when the user logs in, regardless of which machine they log in. So if today they're on machine A, they log in, that profile file gets mounted, they make all their changes, they log off, it gets dismounted and stays where it is. And then tomorrow when they log in on machine B. They see exactly the same thing and they're none the wiser that they've changed underlying host VMs. Okay, perfect. Yeah, that's that's a really important distinction to to sort of the modernization of this as well. One other question is maybe a bit of an aside, not necessarily in the realm of your expertise or maybe it is. But one of my favorite pieces of software ever was Softricity, I don't know if you remember these guys. It was what what got acquired and became App V. So I wonder like is that is that a solution that you you guys have looked at much and speaking of the containerization. It it was really, really effective at containerization of software. Yes, yes. That's that's a that's a good question because now that WVD has been around for about a year and a half or so. The focus is really shifting over to application containerization, right? And FB was was an example of such a technology. The new technology, the new kid on the block that's very popular is called MSIX App Attach. And the concept behind MSIX App Attach is first of all, let's talk about MSIX. So, you know, the Windows installer format traditionally has been MSI, right? The Microsoft installer format that everyone is familiar with. A couple of years ago, Microsoft released an MSIX format, which is sort of a new container type format that can be used to package up and deliver applications. Great. Now, with Windows 10 version 2004, so about a year ago, last year in April, the the version that came out then. They included a feature called MSIX App Attach and what that does is it lets you take that MSIX package, which is ultimately just a zip file, expand it into a VHD file, mount it to your OS, to your, you know, VM. And then instead of installing the application, attach the application. So it sounds very familiar to how profile containerization works where you're mounting the profile rather than copying it back and forth. Same thing with applications. So with WVD, you now have the ability to take your apps, package them into this MSIX format and then deliver them to users at log on time. Similar to kind of how FB and other technologies work. But this is a native OS level technology that Microsoft is, you know, developed and and is pushing and and is, you know, investing into. So natively at the OS level, you no longer must have applications installed, you can connect to them when the user logs in. So the long-term vision is really having a operating system VM that's very basic and generic, it just has the OS and maybe some common applications, then having the user state, the user profile containerized, sitting in an Azure file share or or a file server. And then the same thing with all the applications, and then you define a user profile and you tell it, okay, here's my user profile. And here is my applications that I want the user to get, they log into a blank VM that's just patched to the latest version of Windows. And everything gets attached while they're logging in, and lo and behold, they have their personalized session. But the image itself does not need to have any of that software loaded onto it. Right, yeah. So this is sort of the the perfect sort of panacea of what we had envisioned back then was a strated layers at everything, right? So the the OS is completely segregated from the applications, completely segregated from the users. But you kind of magically merge them together. So it looks like that has really sort of come to the fore. Microsoft has certainly done their work on this. That's awesome. One other piece as a former Citrix administrator, I have some serious battle scars around that I wanted to get your opinion on and how you're managing this in in sort of wide scale multi-tenant environments with a lot of variability on the user end. Is what's your approach around printers? Because printers are basically like the the most hated portions of for most IT people in general. But certainly for people that run a multi-user system, it is it is a created a lot of battle scars for us over the years. Yeah, no, that's that's definitely a a source of battle scars for many. You know, printers or other peripherals, right, you know, if you have scanners or or a USB security keys or or what what have you, those are all very challenging in virtual desktop environments. For printers, there's a number of things that can be done. So we typically recommend or, you know, when when we were running the MSP, the thing we would recommend to customers is have a site to site VPN between the customer's office and the cloud. And then have just regular network printers that could be addressed from the data center and then having printer queues defined on a server that could then be mapped to the virtual desktop. So you really bypass any USB printing or any printer redirection. You're just printing natively via the network, except that network is across a VPN on the WAN rather than local. That's the ideal scenario. You know, nothing special about that, works really well, really reliable and and pretty simple to manage. Not always possible because especially if users are working from home, you're not going to put, you know, a VPN firewall at everyone's home to get that going and not everyone has an IP capable printer. That's only feasible when you have an established office location. So in those scenarios, printer redirection through whatever the display protocol or the presentation layer is is generally the fallback option. With VMware, they license technology called thin print that seemed to work pretty well. In the Microsoft stack, RDP has a printer redirection through something called easy print. That's getting better and better, so most printers are just going to work natively through that redirection. So you connect to your RDP or WVD session, if you open up your printers box, it will show you all the printers that are redirected from the local device. The other options that are I think are going to get to be pretty popular, especially in the Microsoft world, is something they just announced is going into GA or has gone into GA, I think this month or last month, it's called Universal Print. And Universal Print is sort of the the cloud print server solution that's integrated into Microsoft 365. So anybody that's got a Microsoft 365 license is automatically licensed for Universal Print. And there you allow Azure as the cloud to to route that traffic. You no longer need VPNs. You no longer need to worry about setting up printer queues, it handles that for you. So that's I think a technology definitely to watch and is going to get to be pretty popular. And I would say those are the three most common methods, there's certainly others, but but those would be the the top three that I've seen. Yeah, perfect. So I assume, I'd be shocked if this weren't the case. But I assume that you guys saw in a serious acceleration of adoption when everyone kind of the beginning of the pandemic as everyone shifted to work from home. You guys have have certainly seen some success in the channel. But I have to imagine you guys were were fielding a lot of inbound once everyone was forced to work from home. Is that that been a catalyst for people looking at services like Nerdeo? It has been definitely catalyst. And I'll I'll tell you a little bit of a story. So so when we launched our first product called Nerdeo for Azure, it was before WVD. So it was RDS based. It was deploying RDS into Azure. And we started working closely with Microsoft WVD engineering team in in integrating WVD when it launches into our product. WVD went GA on September 30th of 2019, right, so well well before the pandemic started. And we released it the same date when GA in our product. And MSPs who've been using us started deploying WVD and primarily shifted from RDS to WVD. However, what we started seeing is a lot of enterprise customers, governments coming to us and saying like we really like this automation, you don't have to script, you have a nice UI, you have auto scaling. We want to use this, but it just doesn't seem like it's made for us. It was multi-tenant, it was really designed for MSPs exclusively. So they were picking up correctly in the architecture of the product that it wasn't for them. And we decided like, hey, let's create another version of the product that's going to cater to enterprise customers. And we built a new product and launched it on March 10th of 2020. And that product was called Nerdy Manager for WVD. And really was designed for very large scale, you know, up to tens of thousands or even hundreds of thousands of users. And March 10th, I'm sure as many of us will remember, was just a couple days before things went rapidly bad in terms of the pandemic and all the lockdowns. And and as you said, we saw just unending demand for organizations wanting to both use our product and have us help them deploy it and and manage it. And we are not a services organization, we don't provide any IT services. That's why we work with the channel, we work with MSPs. But there was just so many that were coming to us that, you know, several of us were just 24x7 all over the world. Helping organizations get on board into WVD, both the new product and our existing MSP product. And in the course of less than a year, we got about 1,000 enterprise customers using the product and deploying WVD. And probably north of 2,000 SMB products through our MSP channel that have deployed WVD environments. So we've seen a tremendous amount of use, which led us also get market feedback and and, you know, add features to the product. So definitely a huge accelerant, as you said, COVID has been with the whole work from home and remote work scenario. Yeah, well, congrats on the success on that. Obviously, being a solution that was in high demand and and certainly served an urgent need as well. So that's a a good intermix of a solution and a and a need. From an unfortunate need, I suppose. But at least you guys were able to to support those groups in in figuring out something. Because as I'm sure you saw, there was a lot of organizations that were somewhat prepared for this and had to scramble to scale. And there was other people that were just completely flat foot as far as getting people some usable solution in order to work from home. Yeah, it's it's good that you guys were able to fill a gap there as well. One other point of pushback that I I hear from people, being sort of the the head of this organization, and I'd love to sort of put this to you and get your feedback on it. Is a lot of people say, well, you know, I kind of looked at Nerdeo, but it's basically, you know, I could probably build that myself. It's it's some fancy power shell scripts and yeah, they have a nice UI. But, you know, do we really need to pay for that, I I think I could do this myself. And I'll caveat this with most of the people that I see say this never actually do that. But, you know, regardless of that, what what would be sort of your pushback against someone that kind of looks at at the Nerdeo platform and says, if I didn't want to learn this solution. I then it's great, but I if I dug in and spent some time on it, I could probably build it up myself. What's your response to that? I love the question. Definitely a fair one to ask and I think need to be always asked for for any solution that, you know, somebody is is considering using. So, we have several products and they really cater to several different audiences. You know, one of the value propositions in our products is definitely that make it easy proposition, where look, learning how to use the Azure calculator, you need a PhD, right? It will take you a long time, you'll need to learn it, you're going to have to always involve your engineering team to to create proposals. We created a wizard, you go through, you get an accurate price, it's all inclusive. It it gives you everything you need, right? So there is there is that aspect of the product is that it makes it easy to work with Microsoft technology that many will consider not to be necessarily MSP friendly or not, you know, not built with an MSP as the end customer, right? So that that's a part of it. And that's probably the most visible thing because it's such a nice UI and makes things so easy. But in addition to that, we have a tremendous amount of proprietary technology and and automation, you know, we have patents on our technology and we have a fairly sizable development team that's, you know, working for years on building functionality, right? So so it's not just the UI. It's also about deployment. So we have deployment automation that adds real value in terms of taking a typical deployment that may take days, weeks or months and brings it down to hours. Now, someone can say, well, I could I could script it and I could do it myself. Sure, I mean, you could. But will it be based on several thousand deployment experience? Will it be best practice? And how long will it take you to do that, and if you value your time, it may be worthwhile to use a third party tool. So that's an that's one another aspect of it. The third aspect would be optimization and auto scale. We've invested very heavily into a robust auto scaling engine that can not only manage the power of VMs in Azure, so power them on and off to save on consumption, but can also provision or create and remove VMs to precisely match the architecture or the infrastructure size to the user demand. And in a consumption based model, which is what all public clouds are, there is a tremendous amount of value there because it it reduces the spend by matching what you need with with what you have. And there isn't anything like that available natively in Azure, period. I mean, can you write something yourself? I mean, sure, if you have, you know, 30 engineers working on it for a few years, you could. But we won't charge you that much, you know, you can just use use our software that that's been battle tested and and used successfully. And the results of our auto scaling are pretty impressive, we have customers that are saving 75% relative to peak on compute and storage. We have customers literally saving hundreds of thousands a month in their Azure bill as a result of using our auto scaling, right? Obviously, very big deployments. So that's one area. Then then there is the whole management concept. So we have a portal that lets you manage multiple Azure customers across all aspects of their IT environment in a MSP friendly multi-tenant way. So you don't have to switch tenants, you don't have to do all these weird things, you know, to manage multiple customers. You have one consolidated UI, and not only can you jump from customer to customer and do everything from something that's very intuitive to an MSP. Rather to like a developer something, which is, you know, what Azure sort of is built for or the UI at least. But also you're able to do cross tenant management features. For example, we have this concept of global images where you can create a desktop image at your MSP level, install all the common apps and then make it available to 10 of your customers. So whenever you make a change, you're not going and managing each customer's individual image within their environment, you're managing one in yours and you're propagating it down into the customer's environment. And again, that's not something that exists outside of our platform. And there's many other examples, but hopefully that gives you an idea of where we add the value. Yeah, the the metering and the smart management of the resource consumption, I think is a is a huge one. And then as you said, the orchestration and and single paint of glass for multi-tenant management is is a big advantage as well. Okay, so yeah, I appreciate that. This is great. We're running a bit long here, so Vadim, I really appreciate your time. Is there anything we haven't touched on that you want to squeak in before we look to wrap up here? I mean, if I can just just do a quick promotional thing. For anyone who's interested in Nerdeo and and trying it out for your MSP. We actually have a special offer in conjunction with Microsoft going on right now. I'm not sure when this is going to air, but it's it's valid until the end of March. And it gives you six months of licensing for Nerdeo free for up to five customers. Unlimited number of users, you know, Microsoft is sponsoring that. And just get get in touch with us, go to getnerdeo.com to find out more. And I'm just trying to let as many people know about it because, you know, they they can benefit if they wanted to try this out and see how it would be to get your customers into Windows virtual desktop with Nerdeo. Yeah, that's a hell of an offer. Good stuff. All right, and any social channels that you would direct people to, we'll link in the show notes to everything, obviously, the website to to follow up with you guys. But if people wanted to reach out to you personally, any socials you would recommend? Sure, I I'm I'm very active on LinkedIn. I'm not sure you'll find me on Facebook. I certainly couldn't find myself. I won't find you. I'm not there either. I'm a little bit on Twitter, but not much. But I am active on on LinkedIn. We do try to to to really release content of pretty much a daily or every other day, I I release a video on new features. So for anyone who wants to learn about WVD and learn about all that Nerdeo has to offer, you know, LinkedIn would be a good place to start. And feel free to reach out to me directly on LinkedIn as I said, I'm pretty active, so I'll get back to you quickly. Cool. All right. Well, thanks for your time, Vadim. This has been excellent. Thanks for having me.