Evolving VDI – Amitabh Sinha, Workspot

This is Textron tv. Hi everybody. Mike Rothman here, general manager of Techron Research with another Textron TV interview.

Uh, I'm pleased to be joined today by Amit Todd Sinha. He is the C e O of Workspot. We're gonna talk about all sorts of, of stuff, but really focusing on how to both securely cost effectively and with, uh, kind of the proper hygiene.

Really bring a lot of remote and disconnected users, uh, into the environment and do it in a way that doesn't, uh, sacrifice any of the, uh, compromises that we've had to make over many years using old technologies like, uh, V D I and potentially remote browser isolation. Amitha, how are you today? Welcome to the show.

Great, great, Mike. Great to talk to you. Yeah.

So why, why don't you kind of introduce work spot a little bit, talk a little bit about your background. I know you have a, a long and industrious career, and I, I, I don't mean long because you have, uh, probably a little bit of gray hair like I do, uh, but your background is, is pretty impressive. So why don't we go through that a little bit?

Yeah, thanks. Um, so Workspot is a V D I platform company. Um, and what that means is for, for customers, um, that are looking to deploy virtual desktop technology for use cases where in compliance use cases, security use cases, consulting contractors, remote employees, um, all of the stuff that is really, really important for customers to get right, right now, um, V d I is one of the, the best technologies to use.

Um, I was at Citrix before workspot, um, and Citrix is one of the leading on-prem B d I vendors, and I used to lead the product and the go-to-market teams there. Um, and, and uh, what we were focused on that at that point, 10 years ago, 12 years ago, was how do we build the most robust on-prem V D I platform possible? And what we're trying to do at Work Spot now is how do we build the, the best V D I platform for the hybrid multi-cloud world that exists now?

Um, and the problems are fundamentally different. Um, and that's, that's what drives us every day. Yeah.

Yeah. So let's dig into that a little bit. So we all know the old technology, you know, you have 50 or a hundred or 300 contractors, you know, you set up a bunch of, uh, basically b d I instances in a server rack, you know, that sits in your data center.

You use technology again, like your old shop, uh, in order to, to make that happen. What, what were some of the issues that, you know, kind of your customers ran into, uh, that really forced them to rethink kind of the architecture of, of how they were, you know, serving up a lot of these, uh, end user computing devices? Yeah, I mean, I think first, not 300 users, but like their companies that have to put 300,000 users on that.

Yeah. Different, different scale model for sure. Um, I think probably the cost and complexity are probably number one and two.

Um, and complexity is just, um, the amount of effort it takes to rack and stack server storage, networking, load balances, brokers, licensing servers. I mean, just, it's an enormous array of, of infrastructure components you need to deploy to run that thing. Um, and most people were able to do it in one, maybe two data centers around the world.

And so you had people all over the world coming back a hundred, 200 milliseconds away into a single data center running in New Jersey. And the user experience was poor for most users, but that was the only way of delivering secure virtual desktops, um, for the last 25 years effectively. Um, and so that's what people have been doing.

Um, now in the modern world, it becomes even more complex because I think before we had one layer of technology, which was from VMware called vSphere, E S X, which standardized what the interface to the rest of the world was. But now as you go forward, there's a w s has its own APIs. Microsoft has its own APIs, Google has its own APIs, um, and they're very different.

Um, and you're also going from a CapEx to an OPEX model. So it's like, do you have the people that you need to actually run things across these four different infrastructures do know how to manage cost across these four different infrastructures, you know how to run it reliably across these four different infrastructures. And it's becoming an incredibly challenging problem for most customers we talked to now.

Yeah. Well, well, you know, what you just kind of talked about was the challenges of being multi-cloud, right? Yes.

And every company of scale kind of says, well, I have one when I can have four, right? Yeah. You know, you know, the minimum three, right?

And we've done a whole mess of research, you know, especially last year that really kind of talked about co uh, you know, larger organizations are open to, uh, alternative cloud providers or maybe distributed cloud providers, uh, because the specific cloud will support their use case or their application better, right? Maybe it's an analytics application that would lead to certain provider, maybe now it's an AI or gen ai, you know, issue, and then there's a provider that does that. Maybe it's just simple WordPress, right?

And not maybe I don't wanna pay for what, you know, Amazon or, or Microsoft is gonna charge me for that. I'll go with one of the, you know, cheaper, uh, issues. But as you just mentioned, you have all those issues with management, you have issues with cost control, uh, you have issues with just, you know, again, skills, right?

Skills become, you know, one of the huge constraints, right. That we're dealing with, uh, there. So yeah, huge, huge problem, uh, on that front.

But again, let's kind of get back into the, the VDI I piece of it, right? 9% availability. That's right.

And for a workload that's eight hours of downtime a a year. So for, uh, the workload like ours, v d i, where reliability can be the most important thing for many customers where anything below four nines is sort of not acceptable. You cannot do it with one cloud, and so you have to have a multicloud strategy.

Yeah. And so that's another reason why people are going towards multicloud with workspot, which sort of then brings in the cost talent, et cetera, problems into the, into the equation. Yeah, you bet.

As, as we like to say, uh, if you're bad at security in one cloud, it's probably not gonna get a lot better when you go to two clouds, right? Exactly. Exactly.

So you, you've got issues on, on that front. So you know, it's not feasible to, you know, continue to do this in your own data centers, right. You know, just racks and stacks or multiple data centers.

I think covid, you, you know, and again, it's hard to remember there was this thing covid, right? And we couldn't get into the office, uh, for, you know, a couple years, uh, that really put stress on a lot of these specific infrastructures, especially if you're large and you're highly regulated, having folks come in from a, you know, kind of a, an unknown environment, you know, on, you know, a device that, that may not necessarily be controlled, uh, to the degree it needs to by the organization becomes a big issue, right? From a compliance and a regulatory standpoint.

So that starts to break the existing model. The good news is you do have, you know, these things called cloud providers that, you know, can provide, and I'm not gonna say boundless scale, but probably more scale than you can consume, uh, if you're a typical, uh, enterprise. So I, is that kind of what precipitated you guys to start up work spot to really solve this problem and, and provide that layer on top of a multi-cloud environment so that customers didn't have to deal with, uh, a lot of those, you know, management issues themselves.

So I think when we started back in 2012, it was primarily to just to simplify v D I for on-prem, because there was only one sort of real public cloud back then a w s Yeah. And then as we started, so we took the V D I platform layer that was run on-prem and moved it to the cloud, but everything else stayed on-prem. Um, the infrastructure.

And then very shortly thereafter, I think within about a year, two years of launching that product, um, Azure and, uh, we started collaborating with Microsoft Azure to launch our product with them. Then just when the pandemic hit, Google came to us and said, Hey, uh, all our customers want something like this. And then we went onto Google, and very recently we started working with Amazon.

So we did start off with a multi-cloud thesis. It became a multi-cloud because customers and partners are pulling us in that direction. That's right.

That's right. So, so tell me a little bit about you. And again, so I have this issue, I wanna start using your stuff.

What does the migration look like? I've got two data centers full of crap that I have to deal with, right? I don't wanna pay the license fees anymore.

I don't wanna, you know, dedicate that amount of hardware. I wanna repurpose it to something a little bit more productive. Um, how do they get started with, with work spot?

What does that process look like? And, and there are two, uh, two use case, two different scenarios. One is where a customer is already in the cloud, um, and five years ago that was less often the case.

Now it's more often the case that customers are already in the cloud and few customers are not. So the customers in the cloud, they've already set up most of the things they need in terms of networking and security and management and so on. And so for us, what we recommend is treat this like a workload that runs in an isolated part of your data center.

Um, give us your Windows image from your PCs, and you get started. And it's fairly efficient at that point, which is you can be up and running in a day or two, um, with a production, uh, with a pilot use case, test it out. And we insist that you did test it out with some users, right?

Because at the end of the day, if, if users are unhappy, the, the project goes nowhere. So test it out with users. Um, we get most of our customers into production in about 45 days.

And then how they scale is up to them because the first I think a thousand, 2000 users is, is relatively easy. Yeah. But if you start going beyond that, then you can't do it within a month or a quarter.

Like, then it spans multiple quarters, and at that point, that's a customer's responsibility. Yep. Yep, yep, yep.

So tell me a little bit about kind of the security issue, because you, you, you know, and that's one of the things you pointed out, that's one of the main drivers for why folks think about, you know, some type of V D I uh, environment. So what have you guys done on the back end to make sure that you don't have cross data pollination or, you know, kind of making sure that, that the tenants are isolated to the degree that they need to be? So isolation is, is a property of most of the clouds that we work with.

So our, our stuff is multi-tenant. The tenants are isolated from each other, um, by sort of design. Yeah.

Uh, and then each one of the three public clouds have isolation of different kinds that isolates their tenants, right? So, um, we wanna honor that to the highest extent possible, right? So, um, our design actually says, Hey, how do we actually make sure that we are, um, fully isolated from each customer's perspective?

Um, and then the last part I think is just, uh, actually two different things. One is doing everything we can on the client to make sure that the client is fully, uh, trusted. Um, so even if you're coming from an untrusted device, you wanna make sure that it's fully patched, et cetera, so that, um, you're not opening yourself up to risk from that client's perspective.

And then the last part is data, which is we collect a lot of, um, thin data, s I E M data, um, that, uh, companies can integrate with their Splunk system, for example, to give them just 360 degrees view of what's happening in their organization. Um, and so, um, I mean, I would say the amount of data that a system generates is, uh, is, has increased every year that we have gone because the more the, the data people see, they want more data from us and we give them more data. 9% of the, the workload on our systems is actually data that we collect for observing what's going on in the whole environment.

Interesting. Alright. And, and in terms of relative costs of doing it yourself, where, where can I say is, is it about saving money?

Is it just about providing a better, you know, service or more secure type of, of utility to all my end user devices that are out there? So I mean, the larger customers tell us they save a lot of, um, um, uh, people resource where they can dedicate those resources to doing other more useful things in life, right? So, um, we have one customer that said, Hey, uh, if I did it myself, I would take one resource for every thousand users, uh, one full-time resource.

And with workspot, I need about a quarter, uh, for every 10,000 users. So it's a dramatic change, but like, we don't, I mean, most people don't want to hear the, um, I need, uh, I can lay off people because I brought this new technology. And so we don't actually pitch that story.

It's more, Hey, it simplifies your life. So two things happen. One is, um, you don't need the skill sets you needed before, so you can manage it with lower skill sets, so it makes it easier to find the people that you need to manage work spot, right?

And then two is why should you have your people in this day and age learn how to rack and stack servers when they can be doing more business, um, um, value things from an IT perspective. And I think the other thing that we are hearing is most, as most of the sort of the new workforce come in, the people who knew how to rack and stack servers and do all the, the difficult stuff on-prem are sort of aging out of the workforce Yeah. And the new people who come in and don't wanna do that anymore.

Right. So it's sort of like you have to build for the generation that's coming in that either doesn't want to or doesn't have the skills to actually do the things that people could do OnPrem before. Yeah.

They, they all wanna become prompt engineers, right. You know, just, That's exactly right. That's exactly right.

Yeah. Play around with generative ai, this, this old, you know, cis admin stuff is for the birds, right? That's right.

That's, we have cloud people to go and, and do that, you know, kind of stuff at, at, at scale. Uh, so that's really interesting. So, great.

So, so h how do you folks get in touch with you? What are the, what's the best way to, you know, kind of get started if, if you do have this problem scaling up your V D I infrastructure, making it more efficient? I mean, just the easiest thing is come to our website, sign up for a demo, um, and that's like we show you the product in action.

Um, I think our, um, our mantra insight has become just talk less, show more, right? Which is, and, and like what we have found is the more we talk, the less impact it has. Like, let me show you the product and then why don't you go use the product?

Because that's the best way to learn that it is simple. Like the more words we use to describe it, the, the, the more it distracts from the core simplicity of the product itself. Yeah.

That, that was always the, the challenge, right? When I was doing product stuff, it was, we want the product to really show, it's like, why am I sitting spending all this time on PowerPoint? Right.

And little boxes and stuff. It's just like, just get in the product and show me what it does. Yeah.

And, and the problem is you're not, again, I come from a security background and most security folks aren't that good at, uh, uh, you know, kind of the user experience and, and making sure that it, uh, that the value is apparent, especially when a lot of that happens on the, uh, back end, you know, in terms of intelligence and analytics and all that, you know, kind of fun stuff. But y you know, being able to, to have a demo that you know, really shows off the right capabilities and then a pretty, uh, straightforward proof of concept com, you know, process that, uh, exactly. That gives folks the ability to do that, I think is great.

So, um, domain, where can folks find you guys work? com. com.

Alright. com. You don't have to worry about this, you know, all this, you know, these vanity, you know, kind of domains.

That's right. That's right. com.

So, I'm at Todd. Uh, thanks so much for being here on text from tv. Tell us a little bit about, again, the evolution of V D I and, and why the cloud is really the best place to do that.

Um, any party shots before we, before we take off? No, this is fun. Thank you.

Alright, Great. Thanks. Uh, and we'll send it back to the studio for our next interview.