 Welcome to theCUBE's live coverage of Snowflake Summit 22 from Las Vegas, Seethra's Forum, Lisa Martin here with Dave Vellante. We're going to unpack some really cool stuff next in the next 10 minutes with you. Karl Perry joins us, the director of product management at Snowflake. He's here to talk about Snowflake's new unistore workloads, how it's driving the next phase of innovation. Welcome to the program. Well, thank you so much for having me, this is awesome. There's a ton of momentum here. I saw the numbers from fiscal 23 Q1, product revenue, 394 million, 85% growth. A lot of customers here, the customer growth is incredible as well. Talk to us about unistore, what is it unpacked and how have the customers been influential in its development? Yeah, so unistore is a way for customers to take their transactional workloads for their enterprise applications and now have them run on or be built on top of Snowflake. And now you have your transactional data along with all of your historical data. So now you have a single unified platform for doing anything you need to do with your data, whether it's transactional single row lookups, we can do that, whether it's analytical data across, again, transactional and historical data in a single query that our customers are super excited about this. So what are hybrid tables? Is that just an extension of external tables? Yeah, that's a great question. So hybrid tables are a new table type that we've added to Snowflake and hybrid tables are really kind of just like another table with a couple of key differences. So number one is that hybrid tables provide fast, fine-grained read and write operations. So when you do something like a select star from customers where customer ID equals like A32, that's going to return extremely fast. But on top of that same data, your transactional data, you can actually perform amazing analytical queries that return extremely fast. And that's what hybrid tables at their core are. So what does this mean for, so you're bringing their world of transaction and analytics together. What does it mean for customers? Walk us through, Carl, an example. Yeah, so it's great. So Adobe is a customer that is looking at using and leveraging hybrid tables today and then more broadly Unistore. And frankly, Adobe has been an amazing customer since they started their journey. Just really quickly, they're in phase three. The first phase was customers had data in Snowflake that they wanted to take advantage of with the Adobe campaign platform. And so what they did is they built a connector, basically, into and being able to access customer data. And then they started to look at, well, this thing's working really well, let's try to leverage Snowflake for all our analytical needs. And so that was kind of phase two. And now phase three is like, look, let's go and reimagine what we can do with the Adobe campaign platform by having both the transactional and analytical data in the same platform so they can really enable their customers to do personalization, add campaign management, understanding the efficacy of those things at a scale that they haven't been able to do before. And prior to this capability, they would what, have to go outside of the Snowflake data cloud and then do something else and then come back in. Exactly, right? So they'd have a transactional system where all of the transactional state for what the customer was doing inside Adobe campaign, setting up all their campaigns and everything. And that would be stored inside a database, right? And then they would need to ensure that that data was moved over to Snowflake for further analytical purposes, right? And you know, you imagine the complexity that our customers have to manage every single day, a separate transactional system, an ETL pipeline to keep that data flowing, and then Snowflake, right? And with Unistore, we really believe that customers will be able to remove that complexity from their lives and have that single platform that really makes their lives easier. I mean, they'll still have a transactional system. Will they not? Or do you see a day where they sort of sunset that? I mean, there's a set of workloads that are not going to be the best choice today for Unistore and Hybrid Tables, right? And so we know that customers will continue to have their own transactional systems, right? And there's lots of transactions that customers rely and have entire applications and systems built around, right? Right now, with Hybrid Tables and Unistore, customers can take those enterprise applications, not consumer-facing applications, and move them over to leverage Snowflake, and then really think about reimagining how they can use their data that's both real-time transactional as well as all the historical data without the need to move things between systems or use a ton of different services. The Adobe example that you just gave seems like, I loved how you described the phases there and they're discovering, it's like peeling the onion and discovering more and more, but what it sounds like is that Snowflake has enabled Adobe to transform part of its business. How is Unistore positioned to be so transformational for your customers? Well, I mean, I think there's a couple of things. So one, they have this level of complexity today for a set of applications that they can completely stop worrying about, right? No need to maintain that separate transactional system for that, again, enterprise application. No need to maintain that ETL pipeline. That's kind of like one step. The next step is, I mean, all your data is in Snowflake, so you can start leveraging that data for insight and action immediately. There's no delay in being able to take advantage of that data, right? And then number three, which I think is the most compelling part is because it's part of Snowflake, you get in the benefit of Snowflake's entire ecosystem, whether it's first-party capabilities like easy to manage and enforce really powerful governance and security policies, right? Being able to take data from the marketplace and actually join it with my real-time transactional data. This is game-changing, and then, most importantly, is the third-party ecosystem of partners who are building all these incredible solutions on top of Snowflake. I can't even begin to manage what they're gonna do with hybrid tables in Unistore. So, Carl, I have to ask you, so I talked to a lot of customers and I talked to a lot of technology companies. Explain, so Snowflake obviously was the first to separate compute from storage and the cloud database, and then tons of investment came into that space. They kind of follow you on. So that's cool, you reached escape velocity, awesome. But a lot of the companies that I talked to were saying, we're converging transaction and analytics. I think Gartner calls it HTAP or something. They came up with a name. Explain the difference between what you're doing and what everybody else is doing and what customer benefits you're delivering. Yeah, so I mean, I think that's a really great question. To use the term that you use HTAP, right? It's an industry-understood term. Really, when people think about HTAP, what that is about is taking your transactional data that you have and enabling you to do fast analytical capabilities on that. That's great, but there are a couple of problems that historical HTAP solutions have suffered from. So number one, that acceleration, that columnar format of data is all in memory. So you're bound by the total amount of memory that you can use to accelerate the queries that you want to. So that's kind of problem one. This is not the approach that Snowflake is taking. Most importantly, it's not just about accelerating queries on transactional data, whether it's a single-row lookup or a complex aggregate. It's about being able to leverage that data within the data cloud, right? I don't want to have a separate data set on a transactional system or an HTAP system that can give me great analytics on transactional data, and then I can't use it with all the other data that I have. It's truly about enabling the transformation with the data cloud and completely taking away silos so that your data, whether it's real-time, whether it's historical, can be treated as a single data set. This is the key thing that is different about Unistore. You can take the power of the data cloud, all of it, all the partners, all the solutions, and all the capabilities we continue to add, and leverage your data in ways that nobody's thought of possible before. Governance is a huge component of that, right? So in the press release, you have this statement. As part of Unistore, Snowflake is introducing hybrid tables. You explained that, which offer fast, single-row operations and allow customers to build transactional business applications directly on Snowflake. That's a little interesting tidbit. So you expect customers are going to build transactional applications inside the data cloud and somewhat minimize the work that is going to be required by their existing transactional databases, correct? Exactly, and I think, so let me say a couple things on this, right? So first of all, there's a class of applications that we'll be able to just build on top of hybrid tables and run on Snowflake directly for their transactional needs. I think what's super interesting here though is when you, again, start to talk about all your data. One example that we're going to walk through tomorrow on our talk is being able to do a transaction that updates data in a hybrid table and then updates data in a standard Snowflake table and then either being able to atomically commit or roll back that transaction. This is a transaction that's spanning multiple different table types inside Snowflake and you'll have consistency of either the rollback or the commit. This type of functionality doesn't exist elsewhere and being able to take and build transactional applications with these capabilities we think is transformative. And that's all going to happen inside the Snowflake data cloud with all the capabilities. And it's not like what you're doing with Dell and Pure. It's nice, but it's read only. You can't add and delete and do all this stuff. This is native, first class citizen inside the data cloud. Just like other table types, you'll be able to take on and leverage the power of the data cloud as a normal table that you'd be able to use elsewhere. Got to ask you, your energy and the way that you're talking about this is fantastic, the transformation that it's going to be, how central it is to the product innovation since Snowflake is coming out with. What's been the feedback from customers as there's so many thousands of folks here today? Keynote was standing in your room only. There was an overflow. What are you carrying on the floor here? Well, I mean, I think it was funny in the talk when I announced that primary keys are going to be required and enforced and we got a standing ovation. I was like, wow, I didn't expect people to be so excited about primary key enforcement. I mean, what's been amazing, both about the private preview and the feedback we're getting there and then some of the early feedback we're getting from customers is that they want to understand and really think about like, wait, I can use Snowflake for all of this now? And honestly, I think that people are kind of like, but wait, what would I do if I could have those applications running on Snowflake and not have to worry about multiple systems? Wait, I can combine it with all my historical data and anything that's in the data cloud, like what can I do is the question they're asking? And I think that this is the most fascinating thing. Customers are going to build things they haven't been able to build before and I'm super excited to see what they do. More specifically, my takeaway is that customers, actually application builders are going to be able to build applications that have data inherent to those apps. I mean, John Furrier years ago said, data is the new development kit. And it never happened, the data stack, if you will, separate from the application development stack, you're bringing those two worlds together. So what do you think the implications are of that? Well, I mean, I think that we're going to dramatically simplify our customers' lives, right? A thing that we focus on at Snowflake is relentless customer innovation so we can make their lives better. So, I mean, frankly, we talk to customers like, wait, I can do all this? Wait, are you sure that I'll be able to do this? And we walk through what we can do and what we can't do. And they really are like, wow, this could just dramatically simplify our lives and wait, what could we do with our data here? And so, I think with the announcement of Unistore and also all the native app stuff that we're announcing today, I think we're really trying to enable customers and app developers there to think about and being able to leverage Snowflake as their transactionalism, the system of source. So, I mean, I'm super excited about this. I came to Snowflake to work on this and I'm like, can't believe we get to talk about it. How does this work? What's the secret sauce behind it? Is it architecture or is it? Yeah, so I mean, I think a big part of it is the architecture that we chose. So, number one, a key product philosophy that we have with Snowflake is we have one product. We don't have many. We don't put the onus of complexity onto our customers. And so, building that into Snowflake is actually really hard. So, underlying hybrid tables, which is the feature that powers Unistore, is a row base storage engine, right? And then data is asynchronously copied over into a columnar format. And what this provides, because it's just another table that's deeply integrated with Snowflake, is the compiler's completely aware of this. So, you can write a query that spans multiple tables and take advantage of it and we'll take all the complexity. Whether it needs to be a fast response to a single row lookup, or it needs to aggregate and scan a ton of data, we'll make sure that we choose the right thing and provide you with the best performance that we can offer. We built that intelligence inside of that. Completely built in and amazing, but provided in a very simple fashion. You said you came to Snowflake to do this? Yes, yes. How long ago was that? I came here a little over a year and a half. Okay. And they started working on this obviously beforehand, just envisioning it. Yep, right? Yeah. I mean, this is absolutely incredible. I have been working on this now for a year and a half. Some of the team members have been working on it for more. And it's incredible to finally be able to talk to customers and everybody about it, and for them to tell us what they're trying to do. I've already talked to a bunch of customers, like, well, wait, I could do this or this. What about this scenario? And it's awesome to hear their requirements, right? The thing that's been most amazing, and you'll hear it in the talk tomorrow with Adobe, who's been a great customer, is like, customers give us insanely hard requirements. And what I love about this company is not, well, you know, it's easier to do it this way. It's like, no, how can we actually make their life easier? And so we really focus on doing that with Snowflake. And that's one of the things Frank talked about this morning with that mission alignment being critical there. So it's in private preview now. When can folks expect to get their hands on it? Well, we don't have a date right now we're talking about, but you can go sign up to be notified of the public preview when we get there. I think it's like snowflake.com slash tri dash Unistore, but we'll publish that later. And if you're interested in the private preview, talk to your account team and we'll see if you can get in. Carl, thank you so much for joining Dave and me in an action-packed 15 minutes talking about the power of Unistore, what it's going to enable organizations to do. And it sounds like there's your top on the surface. There's so much more innovation that's to come. You're going to have to come back. Yes, that sounds awesome. Thank you so much. Our pleasure. Per Carl and Dave Vellante, I'm Lisa Martin. You're watching theCUBE's live coverage of Snowflake Summit 22 from the show floor in Las Vegas. We're going to be right back with our next guest.