 From around the globe, it's theCUBE with digital coverage of AWS re-invent 2020, sponsored by Intel, AWS, and our community partners. Welcome back to theCUBE's ongoing coverage of this year's AWS re-invent. You know, normally we'd be in the middle of the Sands Convention Center. We have two sets and 50,000 of our closest friends. We'd be geeking out on cloud. Seems like a long time ago, but the show must go on and it does. Joe Duffy is here. He's the co-founder and CEO of Pulumi. And Justin Fitzhugh is the vice president of engineering for cloud engineering for Snowflake. Welcome gentlemen, good to see you. Thank you. It's great to be here. Joe, I love what you guys are doing, you know, leading your customers to the cloud and really attacking that IT labor problem that we've dealt with for years and years by playing a role and transforming what I would say is IT ops into cloud ops with programmable infrastructure practices. So take a moment to tell us why did you and your co-founder start the company? How you got it off the ground? People are always interested in how you got it funded. You got a couple of Seattle VCs, Madrona and Tola involved, NEA just got involved. So congrats on that. What's the story of your company? Yeah, so my background and my co-founder Eric's background, you know, we spent multiple decades at Microsoft just really obsessing over developer platforms and productivity and trying to make, you know, developers' lives as productive as possible, you know, help them harness the power of software to create, you know, innovative new applications and really spent, you know, time on technologies like Visual Studio and .NET and, you know, it really struck us that the cloud is changing everything about how we develop software. And yet from our perspective, coming from developer land, it had almost changed nothing. You know, most of our customers were still, you know, developing software like they did 15 years ago where it was a typical NTR application. They'd kind of write the code and then go to their IT team and say, hey, we need to run this somewhere. Can you provision a few virtual machines? Can you provision, you know, maybe a database or two? And so, and then we went and talked to, you know, infrastructure teams and found out, hey, you know, folks are really toiling away with tools that are pale in comparison when it comes to the productivity that we were accustomed to on the developer side. And then frequently we heard from leaders that there were silos between the organizations. They couldn't build things quickly enough. They couldn't move quickly enough and cloud native and the new public cloud capabilities just really were pushing on that. And really, you know, but the most innovative companies we kept hearing were the ones who figured this out, who really figured out how to move faster in the cloud. Companies like Snowflake really are leveraging the cloud to transform entire businesses. You look at Uber, Lyft, Airbnb, these companies that really harness the cloud to not just from a technical productivity standpoint but really transform the business. So that was the opportunity that we saw with Columi was let's take a step back. We call this cloud engineering. Let's imagine a world where every developer is a cloud developer and infrastructure teams are enabling that new way of building. Great, so you mentioned cloud engineering. Now Justin, you've done a bit of, bit of cloud engineering yourself better than your day. You know, the, the cube has been following Snowflake very closely since it launched really mid last decade. And we've covered your novel architectural approach and your cloud only mantra. Talk about that and have there been any changes in how you're thinking about cloud adoption and how that's, as that's increased and you've seen new use cases emerge? Yeah, so I think, you know obviously Snowflake was built on the foundation of cloud first and in fact cloud only our only platform our only infrastructure is based on the cloud. But, you know, for us it was absolutely key on how do you develop a platform and a product that's completely elastically scalable and really allows for kind of the pay per use and pay per consumption model. We didn't really, it would be very difficult for us to offer this and to offer a product in this way. And if you start to think about kind of from a cloud engineering perspective we don't have the typical network engineers the typical data center engineers that you might have seen previously. Instead, we're shifting our model and what we do in cloud engineering away from kind of an operations model or even DevOps model towards a software engineering model. And I think that's the, that's the big shift to cloud engineering is that we're looking to hire and we're building a team of software engineers to build systems and platforms and tooling to have the system self-manage as much as possible. Any changes to our infrastructure that we look at any changes in our platform are all through commits and deployed via pipelines as opposed to having operators log on and make these changes. And so that's the shift that I think we're seeing and that's to kind of match the overall snowflake model of cloud first and where the product is. Yeah, and like you said in cloud only, Justin you use Pulumi in your own engineering and also in your product externally is that correct and how so? Yeah, we actually use it in specifically and in our platform in order to kind of deploy to manage and just operate kind of our overall cloud infrastructure. We specifically use it more to focus on the Kubernetes and containerization side of things but that use case is kind of rapidly expanding across the organization. So I'm curious what you guys are seeing in the marketplace, Joe. You know, thinking about cloud broadly what's the impact that you're seeing on businesses? Who are the big players that you see out there? Maybe you could talk about some of the differentiation that you've noticed. Yeah, I think this notion of cloud engineering even three and a half years ago when we got started was in its infancy. We definitely saw that, hey, the world is moving it's shifting left, as Justin was saying and really people are looking for new ways to empower developers but that empowerment has to come with guardrails, right? And so what we're seeing is oftentimes teams are now modernizing their entire platform their infrastructure platform and they're looking to technologies like Kubernetes to do that but increasingly, you know, AWS, Azure, GCP you know, when we started there weren't any great managed Kubernetes clusters and now today fast forward, you know only three and a half years and many of our customers were using Pulumi to help them get up and running with EKS in AWS for example, you look at a lot of folks transforming on-prem as well. Again, many times adopting Kubernetes is sort of a if they intend to stay on-prem to at least modernize their approach to application and infrastructure delivery and that's where Pulumi really can help. It can be a bridge to, you know, hey from on-prem to the public cloud there are certainly a lot of folks doing great work in this space. You know, I think VMware has really kind of emerged as sort of a vanguard thought leader in this in this space, especially with, you know Heptio and now kind of pivotal joining the story. We see other, you know, great companies like HashiCorp for doing good work in this space and certainly we integrate with a lot of their technologies and you combine those with the public cloud providers there's also a lot of just smaller startups in the space which, you know strikes in my heart. I love supporting the startup ecosystem you know, whether that's Vercel or Netlify or Serverless, you know, really trying to help developers harness more of the cloud. I think that's an emerging trend that we're going to see accelerating in the coming years. Yeah, thank you. You've mentioned a number of interesting emerging tools companies in the ecosystem. I mean, Justin talked about Kubernetes are there other tooling that you're using that might be, you know some of your customers might like to know about? Yeah, I think, so one thing I wanted to actually follow up with what Joe said here is around kind of the multi-cloud nature of what we do is the tools like Pulumi are critical for us to be able to abstract away specific cloud provider APIs and such and so given Snowflake operates on all three major public clouds and offers a seamless experience amongst all three of them we have to have something that abstracts some of that complexity and some of those technical details away and that's why kind of Pulumi made sense in this case and has helped us kind of achieve that cloud neutrality piece. In terms of other tools that you're thinking that we're talking about. I think Pulumi is doing a great job kind of on some of these on some of the kind of that interaction and infrastructure instantiation but we're looking for tooling to kind of look for the overall workflow automation piece and orchestration. So what sits on top of say you're using Ansible using Terraform you may be using Pulumi as well but what kind of orchestrates all these pieces together and how do you kind of build workflow automation and I think there's a lot of companies and technology providers that are starting up in this area to kind of stitch all these pieces together so that you kind of have this seamless kind of workflow across your infrastructure. Got it. So Joe, I'm kind of curious you talked a little bit about your sort of background at Microsoft and you're even at EMC where you're helping people manage LUNs. This is a sort of skill set that is not in high demand today or at least shouldn't be. People really need to transform. I've said that a lot on the queue but maybe talk a little bit about the experiences that you've had in the past that informed the direction that Pulumi is taking and where you see it going specifically. I mean, I've been talking a lot about the next decade of cloud is not going to be the same as the last decade of the cloud. How do you see it? Yeah, I think I recognize a clear trend in with cloud computing. Back, I can't remember 13 years ago maybe 15 years ago when the Azure project started Dave Cutler who actually founded the NT project at Microsoft actually was one of the first engineers that started Azure and he called it a cloud operating system. And I think that vision of, hey, the cloud is the new operating system is something that we're still just chipping away at. And that was a clear trend. Having seen these transformations in the past, the shift from DOS to Windows, from Windows to mobile to client server to now the cloud, every step of the way we always transform the way we build applications. And I think where we're at now is really in the midst of a transition that I think we'll look back. You never know when it's happening, right? But you can always look back in hindsight and see that it did happen. And I think the trend that we're going through now with service meshes and just microservices and serverless is really we're building distributed applications. These cloud native applications are distributed applications. And that was the trend that I recognized. I also recognize another trend, which is, we spent 30 years building great tools, IDEs, test frameworks, sharing and reuse, package managers. We figured out static analysis and how to fix security problems in programming languages that we've got today. Let's not go rebuild all of that. Let's leverage that. And so that's what Eric and I said on day one. Let's stand on the shoulders of giants. Let's leverage all this good work that has come before us. And let's just apply that to the infrastructure domain and really try to smooth things out, give us a new sort of level playing field to build on from here as we go forward. I'm excited that Pulumi gives us that foundation that we can now build on top of. Great. And Justin, of course, we're covering AWS re-invent. You guys, that was kind of your first platform. It's your largest, the largest component of your business. And I've been saying a lot that, you know, the early days of cloud is about infrastructure, S3, EC2, throw in some database, but really there's a new workload that's emerging. And you guys are at the heart of that where people are putting governed data, giving access to that data, making it secure, sharing that data across an ecosystem. So that new workload is really driving new innovation. I wonder how you see that, what you see the next, you know, half a decade or decade looking like in terms of innovation. Yeah, I think it's a valid point, which is it's less about infrastructure and more about the services that you're providing with that infrastructure and what value are you able to add? And so I think that's at Snowflake, the thing that we're really focused on, which is abstract away all these tunes and all these knobs and such and how much memory you have on a specific in a piece of infrastructure or disk or anything like that. So what's the business value and how can we present that business value in a uniform way, regardless of kind of the underlying service provider and maybe to a different class of business users, someone who wants to load data and just do analytics against it. They really don't want to understand what's happening underneath. And I think that's where this cloud engineering piece comes in and what my team is doing is really focused on how do we abstract away the kind of lower level infrastructure and scalability pieces and allow the application developers to develop this application that is providing business value in a transparent and seamless way and in an elastic way, such that we can scale up and down. We can, we have the ability obviously to replicate both within regions and clouds, but also across different clouds. So from a business resiliency and an uptime point of view, that's something that's been really important. And I think also how do we, security is becoming, is obviously a huge importance given the classification and type of data that people are putting within our platform. So how are we able to ensure that there's a pipeline where developers have reviews and commits of any kind of changes going into the system and their arms length away and can be fully audited for various compliance and regulatory needs. And that's something that kind of this software engineering, cloud engineering concept has really helped develop and allowed us to obviously be successful with various different types of industries. Joe, we're almost out of time. I wonder if you could bring us home. I mean, some of the things Justin was talking about, I mean, I definitely see a lot of potential disruption coming from the world of developers. He was talking about consumption models, different than many of the SaaS pricing models. And how do you see it? Developers are kind of the, really the new source of innovation, your final thoughts. Yeah, I think we're democratizing access to the cloud for everybody. I think, it's not just about developers, but it's really all engineers of all backgrounds. It's developers, it's infrastructure engineers, it's operations engineers, it's security engineers. Justin was mentioning compliance and security. These are really critical elements of how we deliver software into the cloud. And so I think what you're going to see is you're going to see a lot of new, compelling experiences built thanks to cloud capabilities. You know, the fact that you've got AI and ML and all these infinitely scalable data services like Snowflake, and just an arms length away that you can use as building blocks in your applications, application developers love that. If we can just empower them to run fast, they will run fast and they'll build great applications. And infrastructure teams and security engineers will be central to enabling that new future. And I think you also see that, you know, infrastructure and cloud services will become accessible to an entirely new audience. You know, kids graduating from college, they understand JavaScript, they understand Python. Now they can really just harness the cloud to build amazing new experiences. So I think we're still, you know, still early days on the transition to the cloud. I know we're many years on this journey, but we've got many, many years, you know, in our future and it's very exciting. Well, thanks you guys. Joe and Justin, I really appreciate it. Congratulations on your respective success. I know as Joe said you got a lot more work to do, but I really appreciate you coming on theCUBE. Awesome, thank you, Dave. You're welcome. All right, so we're here covering re-invent 2020, the virtual edition. Keep it right there for more great content. We're unpacking the cloud and looking to the future. You're watching theCUBE.