(informative tune) >> From San Francisco, it's The Cube. Covering D2 iQ. Brought to you by D2 iQ. (informative tune) >> Hey, welcome back everybody! Jeff Frick here with theCUBE. We're in downtown San Francisco at D2 iQ Headquarters, a beautiful office space here, right downtown. And we're talking about customers' journey to cloud data. We talk about it all the time, you hear about cloud native, everyone's rushing in, Kubernetes is the hottest thing since sliced bread, but the at the end of the day, you actually have to do it and we're really excited to talk to the founder who's been on his own company journey as he's watching his customers' company journeys and really kind of get into it a little bit. So, excited to have Tobi Knaup, he's a co-founder and CTO of D2 iQ. Tobi, great to see you! >> Thanks for having me. >> So, before we jump into the company and where you are now, I want to go back a little bit. I mean, looking through your resume, and your LinkedIn, etc. You're doing it kind of the classic dream-way for a founder. Did the Y Combinator thing, you've been at this for six years, you've changed the company a little bit. So, I wonder if you can just share form a founder's perspective, I think you've gone through four, five rounds of funding, raised a lot of money, 200 plus million dollars. As you sit back now, if you even get a chance, and kind of reflect, what goes through your head? As you've gone through this thing, pretty cool. A lot of people would like this, they think they'd like to be sitting in your seat. (chuckles) What can you share? >> Yeah, it's definitely been, you know, an exciting journey. And it's one that changes all the time. You know, we learned so many things over the years. And when you start out, you create a company, right? A tech company, you have you idea for the product, you have the technology. You know how to do that, right? You know how to iterate that and build it out. But there's many things you don't know as a technical founder with an engineering background, like myself. And so, I always joke with the team internally, this is that, you know, I basically try to fire myself every six months. And what I mean by that, is your role really changes, right? In the very beginning I wrote code and then is tarted managing engineers, when, you know, once you built up the team, then managed engineering managers and then did product and, you know. Nowadays, I spend a lot of time with customers to talk about our vision, you know, where I see the industry going, where things are going, how we fit into the greater picture. So, it's, you know, I think that's a big part of it, it's evolving with the company and, you know, learning the skills and evolving yourself. >> Right. It's just funny cause you think about tech founders and there's some big ones, right? Some big companies out there, to pick on Zuckerberg's, just to pick on him. But you know, when you start and kind of what your vision and your dream is and what you're coding in that early passion, isn't necessarily where you end up. And as you said, your role in more of a leadership position now, more of a guidance and setting strategy in communicating with the market, communicating with customers has changed. Has that been enjoyable for you, do you, you know, kind of enjoy more the, I don't want to say the elder states when you're a young guy, but more kind of that leadership role? Or just, you know, getting into the weeds and writing some code? >> Yeah. Yeah, what always excites me, is helping customers or helping people solve problems, right? And we do that with technology, in our case, but really it's about solving the problems. And the problems are not always technical problems, right? You know, the software that is at the core of our products, that's been running in production for many years and, you know, in some sense, what we did before we founded the company, when I worked at Airbnb and my co-founders worked at, you know, Airbnb and Twitter, we're still helping companies do those same things today. And so, where we need to help the most sometimes, it's actually on education, right? So, solving those problems. How do you train up, you know, a thousand or 10 thousand internal developers at a large organization, on what are containers, what is container management, cluster management, how does cloud native work? That's often the biggest challenge for folks and, you know, how did they transform their processes internally, how did they become really a cloud native organization. And so, you know, what motivates me is helping people solve problems in, whatever, you know, shape or form. >> Right >> It's funny because it's analogous to what you guys do, in that you got an open-source core, but people, I think, are often underestimate the degree of difficulty around all the activities beyond just the core software. >> Mm-hmm. >> Whether, as you said, it's training, it's implementation it's integration, it's best practices, it's support, it's connecting all these things together and staying on top of it. So, I think, you know, you're in a great position because it's not the software. That's not the hard part, that's arguably, the easy part. So, as you've watched people, you know, deal with this crazy acceleration of change in our industry and this rapid move to cloud native, you know, spawned by the success of the public clouds, you know, how do you kind of stay grounded and not jump too fast at the next shiny object, but still stay current, but still, you know, kind of keep to your kneading in terms of your foundation of the company and delivering real value for the customers? >> Yeah. Yeah, I know, it's exactly right. A lot of times, the challenges with adopting open-sourcing enterprise are, for example, around the skills, right? How do you hire a team that can manage that deployment and manage it for many years? Cause once software's introduced in an enterprise, it typically stays for a couple of years, right? And this gets especially challenging when you're using very popular open-source project, right? Because you're competing for those skills with, literally, everybody, right? A lot of folks want to deploy these things. And then, what people forget sometimes too is, so, a lot of the leading open-source projects, in the cloud native space, came out of, you know, big software companies, right? Kubernetes came from Google, Kafka came from LinkedIn, Cassandra from Facebook. And when those companies deploy these systems internally, they have a lot of other supporting infrastructure around it, right? And a lot of that is centered around day-two operations. Right? How do you monitor these things, how do you do lock management, how do you do do change management, how do you upgrade these things, keep current? So, all of that supporting infrastructure is what an enterprise also needs to develop in order to adopt open-source software and that's a big part of what we do. >> Right. So, I'd love to get your perspective. So, you said, you were at Airbnb, your founders were at Twitter. You know, often people, I think enterprises, fall into the trap of, you know, we want to be like the hyper-scale guys, you know. We want to be like Google or we want to be like Twitter. But they're not. But I'm sure there's a lot of lessons that you learned in watching the hyper-growth of Airbnb and Twitter. What are some of those ones that you can bring and hep enterprises with? What are some of the things that they should be aware of as, not necessarily maybe their sales don't ramp like those other companies, but their operations in some of these new cloud native things do? >> Right, right. Yeah, so, it's actually, you know, when we started the company, the key or one of the drivers was that, you know, we looked at the problems that we solved at Airbnb and Twitter and we realized that those problems are not specific to those two companies or, you know, Silicon Valley tech companies. We realized that most enterprises in the future will have, will be facing those problems. And a core one is really about agility and innovation. Right? Marc Andreessen, one of our early investors, said, "Software is eating the world." he wrote that up many years ago. And so, really what that means is that most enterprises, most companies on the planet, will transform into a software company. With all of that entails, right? With he agility that software brings. And, you know, if they don't do that, their competitors will transform into a software company and disrupt them. So, they need to become software companies. And so, a lot of the existing processes that these existing companies have around IT, don't work in that kind of environment, right? You just can't have a situation where, you know, a developer wants to deploy a new application that, you know, is very, you know, brings a lot of differentiation for the business, but the first thing they need to do in order to deploy that is file a ticket with IT and then someone will get to it in three months, right? That is a lot of waste of time and that's when people surpass you. So, that was one of the key-things we saw at Airbnb and Twitter, right? They were also in that old-school IT approach, where it took many months to deploy something. And deploying some of the software we work with, got that time down to even minutes, right? So it's empowering developers, right? And giving them the tools to make them agile so they can be innovative and bring the business forward. >> Right. The other big issue that enterprises have that you probably didn't have in some of those, you know, kind of native startups, is the complexity and the legacy. >> That's right. >> Right? So you've got all this old stuff that may or may not make any sense to redeploy, you've got stuff (laughing) stuff running in data centers, stuff running on public clouds, everybody wants to get the hyper-cloud to have a single point of view. So, it's a very different challenge when you're in the enterprises. What are you seeing, how are you helping them kind of navigate through that? >> Yeah, yeah. So, one of the first thongs we did actually, so, you know, most of our products are sort of open-core products. They have a lot of open-source at the center, but then, you know, we add enterprise components around that. Typically the first thing that shows up is around security, right? Putting the right access controls in place, making sure the traffic is encrypted. So, that's one of the first things. And then often, the companies we work with, are in a regulated environment, right? Banks, healthcare companies. So, we help them meet those requirements as well and often times that means, you know, adding features around the open-source products to get them to that. >> Right. So, like you said, the world has changed even in the six or seven years you've been at this. The, you know, containers, depending who you talk to, were around, not quite so hot. Docker's hot, Kubernetes is hot. But one of the big changes that's coming now, looking forward, is IOT and EDGE. So, you know, you just mentioned security, from the security point of view, you know, now you're tax services increased dramatically, we've done some work with Forescout and their secret sauce and they just put a sniffer on your network and find the hundreds and hundreds of devices (laughs)-- >> Yeah. >> That you don't even know are on your network. So do you look forward to kind of the opportunity and the challenges of IOT supported by 5G? What's that do for your business, where do you see opportunities, how are you going to address that? >> Yeah, so, I think IOT is really one of those big mega-trends that's going to transform a lot of things and create all kinds of new business models. And, really, what IOT is for me at the core, it's all around data, right? You have all these devices producing data, whether those are, you know, sensors in a factory in a production line, or those have, you know, cars on the road that send telemetry data in real time. IOT has been, you know, a big opportunity for us. We work with multiple customers that are in the space. And, you know, one fundamental problem with it is that, with IOT, a lot of the data that organizations need to process, are now, all of a sudden generated at the EDGE of the network, right? This wasn't the case many years for enterprises, right? Most of the data was generated, you know, at HQ or in some internal system, not at the EDGE of the network. And what always happens is when, with large-volume data is, compute generally moves where the data is and not the other way around. So, for many of these deployments, it's not efficient to move all that data from those IT devices to a central-cloud location or data-center location. So, those companies need to find ways to process data at the EDGE. That's a big part of what we're helping them with, it's automating real-time data services and machine-learning services, at the EDGE, where the EDGE can be, you know, factories all around the world, it could be cruise ships, it could be other types of locations where working with customers. And so, essentially what we're doing is we're bringing the automation that people are used to from the public cloud to the EDGE. So, you know, with the click of a button or a single command you can install a database or a machine-learning system or a message queue at all those EDGE locations. And then, it's not just that stuff is being deployed at the EDGE, I think the, you know, the standard type of infrastructure-mix, for most enterprises, is a hybrid one. I think most organizations will run a mix of EDGE, their data centers and typically multiple public cloud providers. And so, they really need a platform where they can manage applications across all of those environments and well, that's big value that our products bring. >> Yeah. I was at a talk the other day with a senior exec, formerly from Intel, and they thought that it's going to level out at probably 50-50, you know, kind of cloud-based versus on-prem. And that's just going to be the way it is cause it's just some workloads you just can't move. So, exciting stuff, so, what as you... I can't believe we're coming to the end of 2019, which is amazing to me. As you look forward to 2020 and beyond, what are some of your top priorities? >> Yeah, so, one of my top priorities is really, around machine-learning. I think machine-learning is one of these things that, you know, it's really a general-purpose tool. It's like a hammer, you can solve a lot of problems with it. And, you know, besides doing infrastructure and large-scale infrastructure, machine-learning has, you know, always been sort of my second baby. Did a lot of work during grad-school and at Airbnb. And so, we're seeing more and more customers adopt machine-learning to do all kinds of interesting, you know, problems like predictive maintenance in a factory where, you know, every minute of downtime costs a lot of money. But, machine-learning is such a new space, that a lot of the best practices that we know from software engineering and from running software into production, those same things don't always exist in machine-learning. And so, what I am looking at is, you know, what can we take from what we learned running production software, what can we take and move over to machine-learning to help people run these models in production and you know, where can we deploy machine-learning in our products too, internally, to make them smarter and automate them even more. >> That's interesting because the machine-learning and AI, you know, there's kind of the tools and stuff, and then there's the application of the tools. And we're seeing a lot of activity around, you know, people using ML in a specific application to drive better performances. As you just said,-- >> Mm-hmm. >> You could do it internally. >> Do you see an open-source play in machine-learning, in AI? Do you see, you know, kind of open-source algorithms? Do you see, you know, a lot of kind of open-source ecosystem develop around some of this stuff? So, just like I don't have time to learn data science, I won't necessarily have to have my own algorithms. How do you see that,-- >> Yeah. >> You know, kind of open-source meets AI and ML, of all things? >> Yeah. It's a space I think about a lot and what's really great, I think is that we're seeing a lot of the open-source, you know, best-practice that we know from software, actually, move over to machine-learning. I think it's interesting, right? Deep-learning is all the rage right now, everybody wants to do deep-learning, deep-learning networks. The theory behind deep-networks is actually, you know, pretty old. It's from the '70s and 80's. But for a long time, we dint have that much, enough compute-power to really use deep-learning in a meaningful way. We do have that now, but it's still expensive. So, you know, to get cutting edge results on image recognition or other types of ML problems, you need to spend a lot of money on infrastructure. It's tens of thousands or hundreds of thousands of dollars to train a model. So, it's not accessible to everyone. But, the great news is that, much like in software engineering, we can use these open-source libraries and combine them together and build upon them. There is, you know, we have that same kind of composability in machine-learning, using techniques like transfer-learning. And so, you can actually already see some, you know, open-community hubs spinning up, where people publish models that you can just take, they're pre-trained. You can take them and you know, just adjust them to your particular use case. >> Right. >> So, I think a lot of that is translating over. >> And even though it's expensive today, it's not going to be expensive tomorrow, right? >> Mm-hhm. >> I mean, if you look through the world in a lens, with, you know, the price of compute-store networking asymptotically approaching zero in the not-to-distant future and think about how you attack problems that way, that's a very different approach. And sure enough, I mean, some might argue that Moore's Law's done, but kind of the relentless march of Moore's Law types of performance increase it's not done, it's not necessarily just doubling up of transistors anymore >> Right >> So, I think there's huge opportunity to apply these things a lot of different places. >> Yeah, yeah. Absolutely. >> Can be an exciting future. >> Absolutely! (laughs) >> Tobi, congrats on all your successes! A really fun success story, we continue to like watching the ride and thanks for spending the few minutes with us. >> Thank you very much! >> All right. He's Tobi, I'm Jeff, you're watching The Cube, we're at D2 iQ Headquarters downtown in San Francisco. Thanks for watching, we'll catch you next time! (electric chime)