 Hitesh, I think they're signaling you to get going here. Cool. Yeah, let's get started. So Randy, our fourth panelist here. So hopefully he'll show up. And we'll have him in the panel whenever he shows up. But yeah, let's get started. Informally, we have already started a few minutes ago. My name is Hitesh. I am an EIR at a relatively new venture partnership based out of San Francisco, Silicon Valley Growth Syndicate. We have invested in around 60 different startups over the last year and a half. My background and here is Randy. So a round of applause for Randy. Thanks. All right, well done. Perfect. So yeah, so my background before that, I started as a developer 20 years ago. And over the last few years, went through the cycle a couple of times myself to start a company and been on the early days of being a founder. And now, as I'm a part of this fund, one of my jobs is to create these PCs around what's working, what's not, and within the open stack ecosystem. Just wanted to see what has happened over the last three or four years. And I will just run through a couple of slides real quick. And then we'll jump into introduction and then questions. And then I want to leave enough room towards the end so that all of you can ask questions, too. So yeah, so as I was preparing for this last night, I saw this on my feed. Oh, boy. Sean, this is Sean. And I was really hoping everybody will show up here at 9 AM. So thanks to all of you, also, to be here. It was the MetaCloud party last night. So the Star Trek theme. So you're going to have a party tonight? Yeah. I just had four shots of expresso. I wish I could claim something so mundane. You don't want to know why it was late. No problem. Thanks, thanks for every one of you that showed up here at 9 AM Thursday. Great, thank you very much. Now, just wanted to quickly see what the landscape looks like. And this is some data from Q4 of last year. A lot of startups. These are just the startups where funding has gone in. This is from your slide yesterday, Randy, from your state of the stack, P4. 502 companies, more than 500 companies in the ecosystem. It's certainly getting bigger. And where we are now, if you see, we are headed to some really interesting times, as you mentioned. That's right. But what's really interesting is what comes beyond that, that age where there's stability. And there are some real nice businesses still to be made, right? Absolutely. I think the first generation of startups had a bunch of challenges, including mine. And I think there's a lot of times people haven't seen it. I don't know if you've seen it before, but in certain technology venues, there tends to be first, second, third, fourth generations of startups actually doing stuff. So I think there is a good opening for the Gen 2, Gen 3 open stack startups. Perfect, perfect. Yeah, so from here, if you see where, who are the people who are investing? So these are some of the investors who have done two or more deals. And as you can see, there is a lot of money that has gone in. And now what has happened towards the end of the cycle, these are some of the companies that were acquired recently. So when I was looking at building this thesis, what better way to actually bring some of these companies here and hear from them what worked for them, what are the patterns? Is there a formula or a set of things that you guys did right and you went here? So we have MetaCloud, we have Cloud Scaling, and Joe, I've heard some rumors there. So we have, I don't know if you will deny or... No, I mean, no, we're out to build a business and that's the path we're on. Perfect, yeah, you are on that path and you provide that unique perspective to this panel. Yeah, why did you invite me? I've a lot of cards. No, it was not just based on the number. Why did you? Because you're a nice guy, Joe. Yeah, so this is a fantastic panel and I'll go through some of my questions, but towards the end, I want to leave enough time for all of you, if you have questions, you can just line up here in front of this mic and we'll take those. So to just kick it off, if you can maybe quickly go through your background, just a quick introduction, Joe, you can start. Sure, yeah, my name is Joe Arnold and I'm one of the co-founders of SwissStack and we build a product around OpenStack Swift. You know, my background is an engineer and so I come at it from that perspective and back in the day building network management, infrastructure, cloud management and saw an opportunity to launch a startup and it was something I always wanted to do and have all the bruises of going through that process, coming from the engineering side, product side and having to learn all about starting a business through this journey. So looking forward to sharing it with you all. I'm Randy Beis, I'm formerly the CEO of Cloud Scaling. I'm still getting used to saying I'm at EMC. It's already been six months. It just boils off the tongue, CEO of Cloud Scaling. Cloud Scaling was one of the earliest proponents of OpenStack. We were part of the launch in summer of 2010. A lot of the other startups you heard of weren't even aware of OpenStack at that time and then we built some of the first clouds, built the first Swift cloud outside of Rackspace which actually Joe led while he was working as a cloud scaler. Everything I owe to this guy. And he did a kick-ass job and that was like six months after OpenStack was launched. So I mean, you know, it was rough, rough around the edges. And so I've been in this thing from the very beginning and I've been on the OpenStack Foundation Board of Directors since it was formed. And then my personal history is pre-dates OpenStack. I was one of the big cloud bloggers and pundits and I personally have been a hands-on infrastructure engineer network system storage information security for 25 years now for better or worse. Great, Sean Lynch. Let's see, I got my start in the 90s at City Search of all places, a small ideal lab-based company at the time. Pre-start-up, or pre-IPO. And got into infrastructure engineering there. They had a bunch of fancy turquoise SGI kit that I wanted to play with. So got, you know, I feel like, wow, Cray Linking computer sounds fun. So that's how I got into infrastructure engineering. And then City Search and Ticketmaster merged. I ultimately moved to Ticketmaster, rebuilt that platform with a small team and grew the systems engineering team, infrastructure engineering team there. Ultimately headed up global operations as SVP there. And in 2008, we embarked on a private cloud project there. At the time, we didn't have OpenStack, so we built our own orchestration system that we were using, I think, the open source version of Zen server. But built something that was able to support the third largest e-commerce engine in the world. We were transacting about 9 billion a year on that platform. Ton of tenants learned a tremendous amount. So took that experience and co-founded MetaCloud. Gentlemen, by the name of Steve Curry, who ran global storage operations at Yahoo. And the core cloud engineering team from Ticketmaster ultimately joined me. And so that was the basis for the technical team at MetaCloud. One of the things when we get into business model, and I'm sure we'll talk about things like minimum viable product, it was kind of interesting we were seed funding, funded in August of 2011, and we had Disney up and in full scale production by April. So it was super accelerated, super minimum viable product. And be happy to talk about that more. Yeah, that's perfect. So just wanted to kick it off with those early days. Joe, you talked about those dimly lit rooms and what was the life like? And how did people take you seriously at that time? How did you hire those initial few key hires? I think it goes back to those first early customers that you want to go out and get. You have to work really hard to convince them that they should work with you. And honestly, the type of customers early on could do it themselves. And that's almost a barometer. And then there's customers like, I think that we're all consuming. So we had, we worked with a few telcos early on or service providers early on. And man, they just consume so much of our team. And those weren't necessarily like the good ones to start out with. For us, it was working with some of the web scale guys early on because they really didn't have any other options out there to get a product from. And we just applied people. Maybe it's kind of similar to what you guys did. Or you just kind of showed up on site. And yeah, we just, we had an idea and a structure of the product. And we had the early version, the minimum viable product early on. And then we just worked with them to make it work. And we were lucky enough in the early days to find a customer who's willing to pay us a license plus pay for engineering work. And that was almost a funding in and of itself. And obviously it was hard to find, but that's kind of what you need in order to kickstart and get the company off the ground. Yeah, I think overselling or over-servicing over-servicing is super important. Absolutely. I mean, we were basically Disney's cloud operations team. And they had to, like, whenever they wanted, you're there. And it was good. It made the product much better as a result. Cool. Randy, what about you? I don't even know where to start. I mean, I remember in 2010, spending most of our time at 10 engineers in a room, small, dingy room. You were talking about poorly lit rooms in the middle of Seoul, Korea. You know, where, I mean, look, the air conditioning wasn't very good. The building was an old central office that hadn't been used in 10, 15 years. There were abandoned floors and we were in Makdong, which is in the middle of nowhere. You know, and even before that, it was really crazy because, you know, we had inked two multi-million dollar deals with internet and Korea Telecom simultaneously and we only had three employees. So I was using my personal network to basically shotgun everybody together and, like, whipping up employment contracts while getting on planes to fly to Korea. I get to Korea and they told me within three days I got to present to the entire Korea Telecom executive team. And so 20 people are supposed to be there, 75 show up. And I'm, like, you know, basically have an interpreter and I'm going through and walking them through sort of the early stuff on, like, what is cloud? 2010, right? Summer or spring, late spring 2010. And so there's a lot of crazy stuff like that. In terms of sort of customer learnings, I don't know what to say. I mean, it was good to bend over backwards for those early customers that were in our professional services era, but it was also really bad, right? I mean, we chewed up an entire team in Korea. I mean, we just lost the team because it was a year of basically sitting in those hot boxes and it was tough. And Joe is running another team in the US for the other project. And, you know, with the product customers, it was actually easier when we made the transition to being a product company. We would lose the customer, but we wouldn't chew up the teams because it wasn't the meat cloud to keep it all going, right? And, you know, so you can kind of do that a little bit. And, but the thing that we found personally really tough is that we were funded as a software company. And we were taking on the mission of really delivering almost more like a hardware experience. And the problem was is that customers all wanted white boxes from SuperMicro and Quanta and all that, they wanted to reduce their costs, but those things that actually don't maintain themselves, it's not like from Dell, when you get like a particular model of box, all the boxes are the same or pretty close. And there's a lot more variability in the white box supply chain. And so we took on kind of managing that. And so our kind of cost structures and business was more like a hardware appliance business, even though we were funded like a software company. You're bringing up a good point. And I hope you don't mind. I'm gonna diverge a little bit from the actual question because I think the most interesting thing with this panel and the space is the variety of business models. I think that's the really interesting piece here. And I think if you look at the OpenStack space amongst this group and the others, that differentiation of business model, I think was key. Because I mean, realistically, we all had great teams. We all had fantastic engineers. We all had great technology. And we were all using OpenStack. So there wasn't that much differentiation there. I mean, really. So I think if you look at where the differentiation actually existed, it was delivering OpenStack in an appliance like Nebula or delivering OpenStack as software, like maybe a piston cloud or delivering OpenStack as software in a highly opinionated manner. I know Randy said a lot about making your private cloud look and feel like the public cloud, right? So that's a very kind of opinionated play. And I think you're obviously delivering software. So we were delivering OpenStack as a service in a very SaaS-like way. And honestly, we didn't start with OpenStack. I mean, these guys got into the OpenStack space before we did. We set out to build a IaaS as a service business. We looked at Amazon, namely EC2. We sort of broke it up into components and said, well, what are they doing well and what are they not doing well? They're selling cloud as a service. And that's really differentiated. And there's a lot of value there. But they're also selling access to x86 equipment, which isn't special at all, right? And so we thought, gosh, can you just take the cloud as a service piece and just deliver that anywhere? Behind the corporate firewall, co-located on the public side, on the private side, drop the public private monitors and just deliver cloud as a service somewhere. And that's really where, so we actually started with CloudStack for like two weeks and realized that it was just an awful, awful, awful stack and moved to OpenStack pretty quickly. But I think that's the interesting point. Like I'd love to talk about, we had a reoccurring revenue stream. We have that kind of annuity effect with a SaaS-like business. And I think that's what amazed me most. I don't know what you guys think, but the differentiation in business models I thought was fascinating with the early guys. Because we were very different. We were. And I think we could take on a product oriented approach because we didn't try to do everything. We were focused on one project in the OpenStack ecosystem. We could take that independent of the rest of OpenStack. And so it was storage and we could go and deliver just the storage stack. And for that, a product worked really well because we didn't have to deal with as many integrations. And then the integrations we did run up into, we could bolt on directly to our own product, which we could have a lot of control over. And so that's why I think it worked for us because it was more self-contained. Maybe not necessarily less complex, but more self-contained. And we could have more control over it. But when we talk about fundraising, one thing to realize is fundraising is wildly different. You can all be OpenStack companies, but when you're delivering a capital intensive business like an appliance or pure software or SaaS, the makeup of your investors are dramatically different. What we as investors look at, your revenue mix from services or from product, the kind of multiples that you will put on services is much lower. And the kind of investors you can attract and will actually fit really well with what you are doing are different. So what did you do so early on, and then those initial two sort of mid-stage, that initial round of funding with Jerry and others and later on, what kind of things did you do? How did you, what were the challenges like OpenStack? At that time, the perception was very different. Today, it's a very different conversation, but what did you go through there? Early on, we made a lot of mistakes, right? So when we first went out to do fundraising, it took a lot of guidance from folks to that we trusted who had gone through this experience lots of times to set us straight. When we first started, we were trying to raise way too much money, way too complex of a structure when we went out, and we just couldn't get that done. Once we got maybe a little bit more realistic with our fundraising approach, then found investors who had the same belief in where the market was going as we did. We both, Sean and I both share an investor named Ryan Floyd. He's actually be in the next session here, so stick around and listen to him. But getting that investor on board who believed in us and the market and the team was great. We had a lot of things wrong in the beginning and we had investors just flush us out because oh, your pricing model is weird or your delivery mechanism is not something I understand. And they're like, by the way, they were right and we changed our model and we changed our pricing structure in the form factor of our product. So it's not like they're wrong, but it was important to find that investor who... I would just add to that. It's so critical to get the right investor at the right phase. So Ryan Floyd is probably the best early stage investor that I know of in the Valley. Hands down. Even if he doesn't invest in you, he's going to... He will roll up his sleeves and do a tremendous amount of work, but stay out of your way and give you autonomy. But then, I don't know if you guys know, Mahat Ibrahim, she was a series B investor. That's a great series B investor. Yeah, Mahat is an example of someone who looked at us and she was critiquing our business in the early stages. And she was exactly right on where we ended up. But because we weren't doing it that way, she's like, no, not right now. It's really important to have the right investor type at the right stage. And even when you're going out and pitching different investors, being able to pitch to those types of people who are going to give you that feedback is great and more possible. I would say don't worry about logo with the VCs. I mean, don't look out to go be Sequoia-funded or it's about the partner. Think about who you want on your board. That's the most important thing, really. So what was the point where you knew you were ready to go to Ryan and what was that transition point? We had some revenue, some customer traction. And that was enough of a proof point. And we had enough of the larger deals that were either in a POC phase, a paid POC phase that showed some credibility. And that's when Ryan stepped up. Perfect. Yeah, let's talk about those, during that phase, when you were growing, at that point, how you were, for all of you, how were you competing with where you are now? These big guys, how were you getting that business and what kind of things were you doing during that growth phase there? I don't think we were competing with each other, I mean, at all, the market's so big. I mean, it was very, very... Our number one competition was DIY all the time. I lost half my deals to DIY. I think we encountered each other once that I know of. Pissed in three times, never let know times. What about Cisco's and EMC's? None of them were in the market at the time. Yeah, I mean, we were the only guys in the market. It was basically us in VMware. Yeah, I mean, I didn't start to see Red Hat and Marantis until last year. And then they showed pretty heavily, but up until then it was just startups and I saw them very rarely. Most of the deals I lost were, like I said, DIY people were convinced that OpenStack was easy and they thought that they would just download it and sprinkle it in their data center and then poof, you know, they'd have a magical cloud. I gotta tell you, I saw that a million times. I know both of you guys did too. I loved it too, because everyone thinks OpenStack is a product and I can't believe people still think that. I mean, it's not a product. It's a series of raw bits to build a product. Part of it is mistakes on the part of the foundation and the community. In the early days you'd have, you know, the website would say cloud operating system, of course it's non-operating system and the evangelist would be out talking about just download it and get it up and run it in your data center and it's like, no, it's not a product. We loved it, right? So because what we did is we had a product, a commercial product and then we also taught people how to roll it out and deploy it by hand. We did both. We didn't, we were not a services company but what we did was look, if you want to roll it up and do it yourself that's actually demonstrating our value to us and by the way, if you are gonna go run it by yourself then super, you're gonna be a contributor in that community and we're gonna get code from you because you're gonna have to in order to. I had to agree with you. It was also really easy when you follow a DIY because you can show a model, right? You can say, look, I know you tried DIY. All right, you know it takes two or three people to curate upstream to look at what's coming. The best customers were the people who had done DIY but the problem is that that takes 18 months which is a, for a start up trying to get traction is a little bit of a death knot. You guys, I don't know about you but you definitely started a way after me and I started early and it was effing hard to close customers. I remember I lost Symantec and there are big, big OB sector point now and I told Vijay at the time, the architect, I said, you're gonna run into all these problems and I just laid them all out by hand. He's like, yeah, yeah, it'll be no problem. I'm like, you're gonna have to hire like 50 engineers to like get it to where you want it to be and he's like, I don't think it'll be a big deal. I'm finished, I'm finished, I'm finished. And then nine months later, I go back and he's like, Randy, you are right. You're right about everything. I had to do all this stuff and I'm listening to him tell me how he rebuilt my product inside his business. I'm looking at him and I'm going, you're like an antivirus and security company. Why the hell are you building your own product to build a cloud? Like it blows my mind still. Yeah, and I would just say that, how much of our time did we spend educating people about OpenStack? All of it? Yeah, and so now the sales cycle, it used to be probably a 10 month sales cycle for all of us and I saw it compressed to six months, three months, two months. Now it's pretty immediate. Yes, definitely. A lot of that was just training around, this is why OpenStack. Just really quick, for all of you if you have questions, if you can start lining up here and then we'll start taking those questions. Yeah, for me, for us, the sales cycle was shortened not by more education on OpenStack. It was about more focus on use cases. And if one thing that I wish we had done sooner, it was figure out what those very, very specific use cases are and go after them sooner. So, okay, storage, let's be a backup target. How about that? You go into these OpenStack environments and people are doing this. POC, they're having to get their engineers, their developers, they're having to get so many people on board to start using the system that it just takes a long while for the adoption to occur. But it's much better to go to those guys and go, hey, you're in charge of this particular workflow. Why don't you connect the dots yourself? That's been a huge shortening for the sales cycle for us. Great. Any questions from the audience? Not, I'll go to the next one. So, from that stage of growth, what sort of things, as they say, it's never too early to start thinking about the end game? I know there's always this goal to build an independent big enough. You don't think about the end game. You can't spend time on it. You're building the business like Joe said. I mean, if you fixate on the end game, like how many people have ever played baseball or a sport or something like this, right? Most of you, if you ever are in the midst of it and you start thinking about whether you're gonna score the goal, what happens? You lose it, right? You can't think about that. You just don't. It's day to day. You're like, what am I gonna do today that is gonna move the business forward? Then the next day, what am I gonna do today to move the business forward? You're not. I wouldn't plan out longer than like a week. Mm-hmm. Seriously. I think my board would have an issue with that. No, I'm not talking about the long range vision and the strategy of the business. I'm talking about the tactical stuff because one day the emergency is that customer. Next day, there's a person who's problematic on the engineering team. The next day, there's something happening with the customer and production, so you can't... But you gotta have milestones for the stages that you are in the fight, right? Come on, Joe. No, no, no, no, no. Of course you have milestones. But you're not marching towards an exit event. So I agree with both of you guys. I think if you focus on running a business really well, by the time you're approached by a serious suitor with a serious opportunity, chances are you're gonna have many millions of dollars in the bank. You're gonna have really meaningful revenue, really meaningful margins, a really meaningful portfolio. And if that's the point where you're at the table talking about a potential sell, you have all the leverage in the world. And you know you can walk away and have a sustainable business, or you can go raise another round really easily. But if you're on fumes, you're in trouble. So that's the difference between an exit at X versus an exit times X times 10, having some cash in the bank. I think we had about 15 million bucks in the bank and meaningful revenues. And a relatively small team with a relatively low burn rate. And we started talking to Cisco. And so it was a good position to be in. That's great. So what were the things that drove you to EMC and Cisco at that point? Why those were there, that was the right thing to do at that time? So I will kind of agree a little bit with Joe in that you always have to, you can never think about selling your company. You don't wanna be there from a mindset point of view but you always have to be willing to do it. You have to realize why you're there. You're not there for any self promotional. I think a lot of founders get caught up in their own egos. But you're there to maximize the return for your board and your investors. And that's why you're there. So anytime you're approached, you have to take it incredibly seriously. We were approached by multiple large cap companies at the same time. And I think from our point of view, Cisco has a tremendous amount of vision. I think if you look at what they're doing with InterCloud, the InterCloud Alliance, their open stack support thus far, our vision was really well aligned. So I think it was a good fit for us. Great. Perfect. How's, so Paul's acquisition, how's life now? It's really great for me. I mean, I have no complaints. My team's seated, they're all happy. They're working on a new product. And I'm kind of out in a CTO role. The way EMC's structured is you've got the core technologies division, which has all the old school stuff. And then you've got the emerging technologies division, which has all the new scale out software that's meant to disrupt the existing business. And then my job is to come in and disrupt that ETD emerging technologies division. So it causes them grief. And so it's like the perfect job for me so far. Great. Sean? It's been great. I mean, Cisco's been super supportive. I mean, if you look at events, just, you know, the event here, we've never had this type of budget to put it on an event here. So super supportive. We have a ton of autonomy. You know, I will say that anytime you're acquired, if you think about it, you were acquired because you were out executing the company that bought you most likely. And so as a small team. So typically, you know, you were thinking one way, you were executing one way and you're likely going to be purchased by someone who is thinking differently or executing differently. So by definition, that means when you get inside, you're going against the grain a little bit, right? And you have to kind of realize you and your team are going to be a catalyst for change. That's what you were purchased to do. So, you know, that's not an easy job. I mean, that takes a lot of endurance. And I'm sure you're experiencing that too at EMC. I mean, it's just, you know. I don't know. They really want the change. It hasn't been too bad. Yeah. No, they want it at Cisco too, but it's work. You have to roll up your sleeves. Do they really want it, Sean? They do. Yeah, absolutely. I mean, look at Lou Tucker. I mean, how invested is Lou and the business in OpenStack? Lou's awesome. Don't, there are other parts of Cisco, though. Yeah. Well, we haven't encountered the other parts yet. Good. We can start with the questions. Go ahead. So my question is on the seed funding. Your opinion on getting the seed funding from a traditional VC versus going to firms like VMware and Brocade or Cisco to get the funding? What's your opinion? Disadvantage, advantages, if I have options. If I would say one thing, I'd say try to avoid seed funding if you can. I mean, the thing that I learned the hard way is that customers are king. If you have customers, even if it's a small amount and a little bit of traction, they'll extrapolate the value of that for you. And as far as the strategic investors go, it's different money than VCs. It's very different money. The way they think about the world is different. Sometimes they want to attach strings to it if you're doing seed funding from them, which I actually don't know that they do seed funding very often. But if they were to do seed funding, you've got little or no leverage. So I would say do it on, start the thing basically from your garage, essentially, until you've got some customers before you do any funding. I did funding, so we were doing $7 million in trailing 12 months revenue as a professional services business when I took $4 million in funding to be a product company and shut the tap off. And that was really, really hard and we did it with no customers. So I think it's different based on the type of company you're starting, B2B versus B2C, what space you're in, technology, et cetera. I will say, and I suspect, I haven't talked to Joe about this, but I bet he'll agree, that for us, having a very sophisticated investor day one was incredibly helpful. I mean, it was someone that, again, I can't say enough good things about Ryan, but having him in day one was super helpful. And also, if he seeds invest, they're gonna do your A round, right? So, I mean, that's sort of a known. And so, they're with you, they have some skin in the game and that's super important. And the other thing about if you get strategic investor, just understand what you're that company and they have an agenda that they're trying to pursue and they want information about different markets and if that's in a space that you're in, they're gonna have access to the financials and hey, is this a hot category or not hot category? And so, they're not necessarily, they're doing it for other reasons than just a financial return. So just make sure you consider that. And always think about, with these strategic investors, there could be some doors that you're closing in the future from market perspective or- If you go get funding from my Juniper, you can't go get- Go to Cisco, so just keep that in mind. And then you're gonna have to see other people say, even opposite stories to everything we just said too, which getting investment, a strategic investment from somebody has opened up channels into that company, but I don't know, that's not necessarily what I would beg. I would go as far as to say, don't ever start with a strategic. I mean, honestly, I think you want an institutional investor early on. It gives you credibility for your subsequent rounds. Absolutely. A tremendous amount of credibility. Yeah, and I think it's more important just to have the right lawyers. Really. Yeah, that's- That's all. Yeah, go ahead. Thank you, this has been extremely invaluable. So my question is on MVP, the minimum viable product, what should we actually have as an MVP before we either go to a customer or go for funding? Because it's kind of a chicken and egg approach, normally I've seen, that you start building something and you go to a customer customer, I want A, B, and C more, then you go back, and by then you have so many competitors. So should we just go with an idea? Is that better, or should we start building a product? There's an idea, and then there's something in between an idea and an MVP. Like, and there's some people I know, like they consider the MVP to be a slide deck that explains how the product will work and they'll go out and they'll run that by dozens and hundreds of customers first, incorporate the feedback you're talking about until they actually have something that's material and they may not write in code then they'll try to go get that funded. That's legitimate. But I can guarantee you that whatever idea that you have for your business, by the time you get done, you'll be doing something else. Like you almost never see, like how many people know where Twitter came from? Twitter came out of a podcasting company called Podio, which died because iTunes started delivering podcasts. And these guys had this little app that they used to coordinate lunch in San Francisco. That's where Twitter came from. I mean, that's just the norm in starting that. You're gonna move left and right and zig and zag. And you guys pivot, we pivoted. No, not really, no. I mean, we started out with a product concept and we, yeah, it wasn't, the first version wasn't very sophisticated, but it was the- What if you look at it from the point of view that you were cloud scaling version 2.0 when you basically figured out that Randy had his head up his arse and wouldn't like focus on Swift and peeled out? Why did it take until 2.0? I'm just kidding. Look, but we had, when we first were talking to customers, we had something that at least got Swift up and running. And that was like our minimum product that we could show customers. And then we just iterated on that from there. But just as a rule, just earlier is better. Don't haul up for six months trying to build something. It'd be kind of the general recommend. I would agree. And don't keep your day job while you're working on your startup. No, don't. That's an awful idea. I've seen so many people do that where they're working part-time on their day job while they're trying to get their startup off the ground. You'll never be successful. You need the gun to your head of that being your only job to make it successful. Yeah, I'm gonna temper that a little bit. I think that's generally true, but what I did in the first go-around, the startup before cloud scaling is I was doing consulting. And I had set up the consulting structure in such a way that it was more like half time, but I was able to put the bacon on table. So that way, I'd have like three full days per week, plus the weekend to actually work on the startup stuff. So you can do other things if you need to, if you really have to bootstrap it, but you can't be working a full-time regular job, for sure. Because that's just, none of those take, nobody here works 40 hours a week, right? So I'd say raise enough before your MVP to at least focus on your startup 100%. That's just my- One quick question. I know that we'll go into what sort of opportunities exist next for OpenStack in our next panel, but just wanted to ask you guys, if you were starting today, like you are, you know that, with your jobs that you have now, you know that other thing, what are those exciting areas if you were starting today around OpenStack? What will those companies be? If I was gonna start over now, I'd do it kind of what Joe did. I'd focus on components. Like we tried to like deliver the distribution, and we eventually did, right? I mean, we were very successful at Walmart, but that's just very hard to do as a startup. You really need a lot of money. So I think it's better to go peel off a piece, whether it's in storage or computer networking and policy management or whatever. I think it's better to peel off a piece. Cool. I'll be a little controversial. I think the trains left the station as far as starting an OpenStack company. I think you're racing dollar for dollar with many, many companies with many, many, many billions of dollars in the bank. So that wasn't the case when we all started. So I'd probably go UpStack. I think there's interesting, I guess maybe I have a bias toward as a service, SaaS type businesses. I think data analytics is a service, that type of business. Cassandra as a service or that type of thing. Sort of cloud agnostic. I think that area of the market hasn't really been as penetrated as it could be. I think there's a ton of opportunity there. I'd plant myself in a vertical. And I would take whatever industry that I knew really well and I would see how I could deliver that with this delivering whatever it is as software on commodity equipment using OpenStack to power that. That's where I think you'd have to figure it out because I think just doing technology is very, very difficult right now because there's so many players out there that are trying to do it just from the pure component. You're doing awesome within BioFarma, for example. Yeah, BioFarma. We're trying to take on BioFarma and media entertainment right now. And that's working because we're putting a lot of energy and focus to it. Go ahead, please. Okay, just one quick, yeah. So everybody talk about these MVPO, of course. So I think we are out of time, unfortunately, but I think we'll be here for some time. And thank you very much, all of you. Thanks for coming. This was awesome. Please, a round of applause for our panelists here. Thank you.