 Hey everyone, welcome back to theCUBE's live coverage of Snowflake Summit 22. We are live at Caesars Forum in Vegas, at least I'm Martin with Dave Vellante. Excited to welcome a VIP, fresh from the keynote stage, the SAP product at Snowflake, Christian Klanerman. Christian, thank you so much for joining us on theCUBE today. Thank you for having me, very exciting. And thanks for bringing your energy. I loved your keynote. I thought, wow, he is really excited about all of the announcements. Jam-packed, and we didn't even get to the entire keynote. Talk to us about, and for the audience, some of the things going on, the product revenue in Q1, fiscal 23, 394 million, 85% growth, lotta momentum at Snowflake, no doubt. So, I think that the punchline is, our innovation is, if anything, gaining speed. We were over the moon excited to share many of these projects with customers and partners, some of these efforts have been going on for multiple years. So, lots of interesting announcements across the board from making the existing workloads faster, but also we announced some new workloads getting into cybersecurity, getting into more transactional workloads with Unistore. So we're very excited. Well, first time being back, this is the fourth summit, but the first time being back since 2019, a tremendous amount has changed for Snowflake and that time, the IPO, the massive growth in customers, the massive growth in customers with over one million in ARR. You talked about one of the things that clearly did not slow down during the last two years as innovation at Snowflake. Yeah, that for sure, like, even though we had a highly in-the-office culture, we did not miss a beat. The moment that we said, hey, let's all start doing Zoom-based calls, we did so, I don't know if you saw the first five minutes of my session in the keynote, we originally talked about summarizing it and no, we're going to spend 40 minutes here, so we did a one-minute clip and whatever gets flashed there. So no, the pace of innovation, I think it's second to none and maybe I'll highlight something that we're very proud of. Snowflake is a single product, a single engine. So if we're making a query performance enhancement, it will help the cybersecurity workload and the high concurrency, low latency workload, and eventually we're starting to see some of those enhancements all the way to Unistore. So we get a lot of leverage out of our investments. What's your favorite announcement? That's like picking children, of course. I think the native applications is the one that looks like, yeah, I don't know about it on the surface, but it has the biggest potential to change everything. Like create an entire ecosystem of solutions for within a company or across companies that I don't know that we know what's possible. Well, I've been saying for a while now that you have this application development stack over here, the database is kind of here, and then you have the analytics and data pipeline stack. There's no separate worlds, but we talk about bringing data and AI and machine intelligence into applications. The only way that that is actually going to move forward is if you bring those worlds together. You're a good example of that happening within a proprietary framework. It's probably going to happen open source organically and you can sort of roll your own. Is that by design or is it just sort of happening? Well, they bring it all into a single platform, obviously by design, because there is so much friction today on making all the pieces work together. Which database do I use for transactions and how do I move data to my analytics system and how do I keep system reference data in sync between the two? So it's complicated and our mission was remove all of this friction from the equation. The open source versus not, the way we think about it is open sourcing, open formats or even open APIs. It's, does it help us deliver the solution that we want for our customer? Does it help us solve their problems? In certain instances, it has done in the past and we've open sourced frameworks. In others, we mentioned at the keynote today the integration of iceberg tables. That's a strong embrace of open technologies. But that does not mean that we want to continue to innovate on our formats. A lot of what you see in the open formats is because snowflakes proprietary innovation. So we have a very clear philosophy around this. Well, like any cloud player, you have to bring open source tools in and make them available for your application developers. But take us through an example of Unistore and specifically how you're embracing transaction data. What's a customer going to actually do? Take a, paint a picture for us. I'm going to give you a very simple use case but I love it because it shows the power of the scenario. Today, when people are ingesting data into Snowflake, you want to do some bookkeeping associating with those loads. So imagine that I have a million files. How many of those files have I loaded? Imagine that one of those loads fail. How do you keep in sync whether the data made or not with your bookkeeping? Today, if you had to do it with a separate transactional database for the bookkeeping and the loading in Snowflake, it is a lot of complexity for you to know what's where. With Unistore, you can just say, I'm going to do the bookkeeping with these new tables called hybrid tables. The loads are transactional and all of this is a single transaction. So for anyone that has dealt with inconsistencies in database world, this is like a godsend. Okay, so my interpretation of that is all about what happens when something goes wrong. Which is a lot of everything about transactions. What happens when goes wrong? And goes wrong doesn't mean failures. Like goes wrong is when you're debiting money from your bank account, not having enough balance, that counts as go wrong and the transactions should be aborted. So yes, transactions are all about conflict management and we're simplifying that in a broader set of use cases. And recovery, so you're in fast recovery. So the business impact of what you're doing is to sort of simplify that process. Is that an easy way to boil it down? Pretty much everything we do is about simplification. Like we've seen organizations at large focusing on wrestling infrastructure as opposed to what are the business problems? For a frank, we reference something that I believe very much in, like which is mission alignment. We are working on helping our customers achieve what they're set out to achieve. Not giving them more technology for them to their goal to become to wrestle the infrastructure. So it's all about ease of use, all about simplification, removal of friction. Just, if I may, so mission alignment. You know, you always hear about technology companies that provide infrastructure or a service and then the customer takes that and monetizes it. Pretty much on their own. The big change that I'm discerning from these announcements is you're talking about directly monetizing and participating in that monetization as a technology partner, but also the marketplace as well. Correct, and I would say in some ways this is not new. This has been happening for the last couple of years with data. Like if you saw our industry data cloud launch as a financial service cloud, it comes with data providers that help you achieve specific outcomes on a specific industry. What we're doing now is saying it's not just data, maybe it's some business logic, maybe it's some machine learning, maybe it's some user interface. So I think we're just turning the knob on collaboration and it's a continuation of what we've been doing. Talk a little bit more about mission alignment. When I heard Frank Slutman talk about that this morning, I always love that when I hear cultural alignment with organizations, but as you just said, it's really about enabling our customers to deliver outcomes to their customers. As the SVP of product, can you talk a little bit about how the customers are influencing the product roadmap, the innovations and the speed with which things are coming out at Snowflake? Yeah, so great question. We have several organizations at Snowflake that are organized by vertical, by industry. So the major sales organization is part of it. The marketplace business development team is organized like that. We have a separate team that provides thought leadership by industry vertical globally, and then even within our solution engineering, there is verticals. So we have a longitudinal view of all the different functions and what do we need to do to achieve a set of use cases in a vertical. And all of those functions are in constant communication with us on versus where the product is seeing an opportunity or could do better for that vertical. So yeah, I can tell you, and obviously we love when there's alignment between those, but that's not always the case. You heard us talk about clean rooms now for some time. Clean rooms are applicable to almost any industry, but it's red hot for media and advertising, third party cookie deprecation and all of that. So we get to see that lens that our innovation is informed by industries. So we're seeing obviously the evolution of Snowflake. We talked about in the key notes today, you guys talked about 2019 and pre-2019 even. It was to me anyway, your first phase was, hey, we got a simpler EDW, you know, we're going to pick that off and put it in the cloud and make it elastic and separate compute from storage, all that kind of cool stuff. And then during the pandemic, it was really your IPO, but also the data cloud concept, you sort of laid that vision out. And now you're talking about application development, monetization, what I call the super cloud, that layer, right? Okay, so talk about- I heard your term in the past, yes. You talk about this, these announcements, how they fit into that larger vision, where are you going? Great question. The notion of the data cloud has not changed one bit. The data cloud thesis is that we want to provide amazing technology for our customers, but also facilitate collaboration and content exchange via our platform. And all that we did today is expand what the content can be. It's not just data or a little helper function, it's entire applications, entire experiences. That is the summing up the impact of our announcements today. That's the end of it. So it's still about the data cloud. So what is impressive to me is you guys couldn't have a company without the hyperscalers, right? It would be a lot different, right? So you built on top of that. And now you have your customers building their own super clouds, I call it. I get a lot of grief for that term. But the big area of criticism I get is, ah, that's just SaaS. And I'm like, no, it's not. No, is everybody public who's announcing stuff? I better be careful, but you have customers that are actually building services, taking their data, their tooling, their proprietary information and putting it on the Snowflake data cloud and building their own clouds. That's different than, that's not multi-cloud, which is, I can run on a different cloud. And it's not, is it SaaS? I mean, it feels like it's something new from your perspective, is it different? I love that you called out that running on all clouds is not what we do. Right. These days, everyone is multi-cloud. You run on a VM or a container and I'm multi-cloud check. No, we have a single platform that does multi-region, multi-cloud, but also cross-region, cross-cloud globally. That is the essence of what we're doing. So it is enabling new capabilities. I've also said, in many respects, that super-cloud hides the underlying complexity. You think about things like exploiting Graviton and a developer doesn't need to worry about that. You're going to worry about that. But at the same time, as you get into the world of application development, some of your developers may want access to some of those cloud primitives. Are you providing both? What's the strategy there? Generally not. In some areas, we bleed through some details that are material, but think of the reality of someone that wants to build a solution. It's really difficult to build an awesome solution in one cloud. You need to do what's the latest instance and is Graviton going to help you or not, all of that. Now do it for another one and then do it for another one. And I can tell you it's really difficult because we go through that exercise. Snowflake porting to a new cloud is somewhere between one and two years of effort and not a small number of people because you're looking at security models and storage models. So that's the value that we give to anyone that wants to build a solution and target customers in all three clouds. I mean, people are still going to do it themselves, but they're going to spend a lot more and they're going to lose their focus on what their real business is. And there'll still be that. I think that DIY market is enormous for you guys, huge opportunity. And there's also the question on what is the cost of that analysis and that effort? And can we amortize it on behalf of all of our customers? We talk about gravitan. We have not talked about the many things that we evaluated that were not better price performance for our customers. That evaluation happened. That value was delivered by not moving there. And when you do it yourself, the curve looks like, okay, hey, we can do it ourselves. We can make it pretty inexpensive. And then the costs are going to decline. But what really happens, like developing a mobile app, you got to maintain it. And then if you don't have the scale and you don't have the engineering resources, you're just the cost are going to continue to go through the roof. A lot of you compare it to mobile apps. Like, I still don't understand why every company that wants to build an app has to build two. You got it. There is no super cloud for the phone. Right. That's sort of our broad vision. Not yet. Not the phone, but the super cloud. Yeah, yeah, absolutely, you get it. This is, and you look out the ecosystem here. I mean, what a difference that you've been pointing this out Lisa from 2019. A lot of buzz, it's all about innovation. You see this at re-invent is like the Super Bowl, obviously, and you see that. And it used to be, oh, how is AWS going to compete with Snowflake and separate compute with stores? That's, I feel like in a large way that's all gone. It's like, okay, how do we rise the whole industry? And that's really where the innovation is. We have an amazing partnership with AWS and they benefit from what we do. Yes, there's some competitive elements, but we're changing so many things, creating so much opportunity that we're more aligned than not. Yeah. Last question for you is continuing on the AWS partnership front. How does a partner like AWS and other partners, how do they fit into the data cloud narrative that you're talking about to customers? I would say that other than the one or two teams that are directly competitive, the rest of their teams are part of the data cloud. Like our relationship with SageMaker as an example is amazing and a lot of what we want to deliver to our customers is choice around machine learning frameworks and tools and they're part of the data cloud. We're working with them on how do you push down computation to avoid getting data out, to reinforce governance? So I would say that, and go look at it, they have a hundred and something teams. So if two teams out of hundreds are the competitive element, we are largely aligned and they're part of the data cloud. Yeah, I mean, your customers consume a lot of compute and storage for AWS and also increasingly Azure and Google. I mean, it's pretty amazing times, Christian. I want to ask you about a couple of terms, one term that came up a couple of times today in Frank's keynote, he said, I'm not going to call it a data mesh. I'm kind of out of respect for the purists, which is cool, I thought. But then you got a customer stand up, Geico said, we're building a data mesh. JPMC is speaking at this event, building a data mesh. And I look at things through that prism and say, okay, data mesh is about decentralization, I'd be curious as to whether or not you tick that box, but it's about building data products, it's about self-service infrastructure and it's about automated computational governance. You are actually ticking a lot of those boxes and Mike, I guess the big one is, are you building a bigger walled garden? But I think you'd say no, it's a giant distributed network, but what do you say to that? We, the latter, giant distributed open cloud and open in the sense that we want anyone to plug in and someone can say, well, but I cannot read your file formats. Sure, you can with what we announced today, but it's not about that. Our APIs are open, we have REST APIs, we have JDBC, ODBC, probably most popular interfaces ever. And we want everyone to be part of it. If anything, there's lots of areas that we would not want to go into ourselves because we want partners and customers to go in there. So no, we're looking at a very broad ecosystem. We win based on the value created on top of the platform. Yeah, it makes total sense to me. I mean, I think the immaculate conception of data mesh might be a purely open source version of Snowflake. I just don't see that happening anytime soon. And so I think you are, I wrote about this, creating a de facto standard. Exactly, and I don't like to get into the terminology there. Is the data mesh or not? No, go look at the concepts. Like people used to say, but Snowflake is not a data lake. Okay, what is a data lake? It's just a pattern, and if you follow the pattern and you can do it, that's fine. Then there's the emotional quasi-religious overlay open versus not. I think that's a choice, not necessarily the concept. I think that's a moving target. I mean, UNIX used to be open. Now the reason why I do think the data mesh conversation is important is because Jamak Tagani, when she defined data mesh, she pointed out, in my view, anyway, the problems of getting value out of data is that you go through these hyper-specialized teams and they're blockers in the organization. And I think you, in many respects, are attacking that. And it's an organizational issue. The insights in the pattern are 100% valid and aligned with what we do, which is that you want some amount of centralization, some amount of decentralization, living in harmony. I have no problem with terminology. And the governance piece is massive, especially the picture's becoming much more clear. Whatever's in the data cloud is a first-class citizen. You give all these wonderful benefits. I mean, the interesting thing what you're doing with Dell and Pure, I asked you that on the analyst's call, it's a start, you know, I mean... And I said it briefly in the keynote this morning. We're publishing a set of standard conformance tests. So any storage system can plug into a data cloud. And by the way, it's based on S3 APIs, another de facto standard. Like it's not a standard, but everyone is emulating that, and we're plugging into that. Yeah, nobody's complaining against S3 API. Oh, it's not an Apache project, we shouldn't. Who cares? Everyone has standardized on that. That's it. Well, we've seen the mistakes of the past with this. I mean, look at Hadoop, right? There was this huge battle between, you know, Cloudera and Hortonworks and Mapbar is proprietary. Oh, Hortonworks is purely open. Cloudera's open core, they're all gone now. I mean, not gone, but they didn't have it right. You know, they got unfocused. I go back to Frank's book. They were trying to do too many of the zoo animals and you can't fund it all. Which for us, it's very important. I can give you, I don't know, 20 announcements or 50 announcements from the conference, but they're all going towards a singular goal. And it's this, do not trade off governance of data with the ability to get value out of data. That's everything we do. And that's critical for every company in every industry these days that has to be a data company to survive, to be competitive, to be able to extract value from data. If data's currency, how do I leverage a tool like Snowflake to be able to extract insights from it that I can act on and create value for my organization? Geico was on stage this morning. Everyone knows Geico and their beloved Gecko. Is there another customer that you have that you think really articulates the value of the data cloud and to Dave's point, how Snowflake is becoming that de facto standard data platform? We had Goldman Sachs on stage as well today. And he mentioned it that people think of Goldman as investment banking and all of that, but no, at the heart of what they do, there's a lot of data and how do they make a better decision? So I think we could run through 20 different examples because your premise is the most important. Everything is a data problem. If it is not a data problem, you're not collecting the right data and getting the insights that you could be getting. These guys are public, right? Adobe, yeah. Right, Adobe's doing it. I don't know if the other one is, I don't want to say, I'll have to ask you off camera, but the other financial firm building a super cloud, right? I call it super cloud. So we're taking advantage of Unistore to bring different data types in and monetize it. That's, to me, that's the future of data. That's been the holy grail, right? We tried to emphasize that this is not a, hey, six months ago we decided to do this. No, this is years in the making, which is why we were so excited to finally share it. Because you don't want to say, three years from now we're going to have something. No, it was the, now we have it, we have it in preview and it's working and it is as close to the holy grail as it gets. Yeah, I mean, look, pressure's on, Christian. Let's face it, enterprise data warehouse failed to live up to the promises. Certainly the data lakes failed to live up, master data management, all that's a Hadoop. All that stuff, there was a lot of hype around that. And a lot of us got really excited, me included, and then customers spent and they were underwhelmed. So, you got to deliver, you say it, you got to do it. Correct, and the other thing is, I would say all of those waves of technology, there was no real better choice. Right, they added value, I wouldn't debate that. You have to give it a shot, like when you bought 20 different appliances and you have all these silos and someone sells you, hey, Hadoop will unify it, it sounds good, just didn't do it. Yeah, and no debate that it brought some value for those that were sophisticated enough to deploy it. I agree. Yeah, but this is a whole different ball game. Well, everything we want to do is democratizing, simplify. Yeah, we could go build something that, I don't know, 10 companies in the world could use. That's not a sweet spot. Like, how do we advance like the state of value generation in the world? That's the scale that we're talking about. It's going to make it easy, accessible for everyone. And govern. Governance is an imperative, it's a law. Yes. But it's not, that's a really difficult challenge to create what I'll call automated or computational governance in a federated manner. That's not trivial. And that's our thesis. Everything we're doing is snowpark. Big announcement today, Python. I've had people tell me, well, but Python should be easy to host the Python runtime. Like, you can do it like, I think in a week. It took us years. Why? Oh, secure. Oh, details. A lot, and Benoit mentioned it, like, securing that is no easy feat. Christian, thank you so much for joining Dave and me. Bringing your energy from the keynote stage to the Cube set, breaking down some of the major announcements that have come out today. There's no doubt that the flywheel of innovation at Snowflake is alive, well, and moving quickly. Innovation is at an all-time high at Snowflake. Thank you for having me. All right, our pleasure, Christian. From our guest, Dave Vellante, Lisa Martin here, live in Las Vegas at Caesars Forum, covering Snowflake Summit 22. We'll be right back with our next guest.