 Okay, welcome back everyone. We're live here in theCUBE in Las Vegas. For AWS re-invent 2021, I'm John Furrier, host of theCUBE. We're in person this year. It's a hybrid event, online, great action going on. I'm with her role of the vice president of AWS analytics at AWS. Great to see you. Thanks for coming on. It's great to be here, John. Thanks for having me again. So you got a really awesome job. You got serverless, you got analytics. You're in the middle of all the action for AWS. What's the big news? What are you guys announcing? What's going on? Yeah, well, it's been an awesome re-invent for us. We've had a number of serverless analytics launches. So Redshift, our petabyte scale data warehouse, EMR for open source analytics. And then we've also had manage streaming for Kafka, go serverless, and then on demand for Kinesis. And then a couple of other big ones. We've got row and cell based security for AWS lake formation. So you can get really fine grain controls over your data lakes. And then asset transactions. So you can actually have inserts, updates, and deletes on data lakes, which is a big step forward. I saw Swami on stage in the keynote. He's actually finishing up now. But even last night I saw him in the hallway we were talking about, it's not just about AI. Of course, he's got the AI title. But AI is the outcome, it's the application of all the data. And in a new architecture, he said on stage just now, like, hey, it's not about the old databases from the 90s. There's multiple data stores now available. And then he's got the unifications, the big trend. And he said something interesting. Governance can be an advantage, not an inhibitor. This is kind of this new horizontally scalable kind of idea that enables the vertical specialization around machine learning to be effective. It's not a new architecture, but it's now becoming more popular and people are realizing it. You can share your thoughts on this whole, not shift, but the acceleration of horizontally scalable and vertically integrated. Yeah, I think the way Swami put it is exactly right. What you want is the right tool for the right job and you want to be able to deliver that to customers you're not compromising on performance or functionality or scale, but then you wanted all of these to be interconnected so they're well integrated. You can stay in your favorite interface and take advantage of other technologies. You can have things like redshift integrated with SageMakers, you get analytics and machine learning. And then in Swami's absolutely right, governance is actually an enabler velocity. Once you've got the right guardrails in place, you can actually set people free because they can innovate. You don't have to be in the way, but you know that your data is protected. It's being used in the way that you expect by the people that you are allowing to use that data. And so it becomes a very powerful way for customers to set data free. And then because things are elastic and serverless, you can really just match capacity with demand. And so as you see spikes in usage, the system can scale out. As those dwindle, they can scale back down. And it just becomes a very efficient way for customers to operate with data at scale. Every year at re-advent, it's always kind of like a pinch me moment. It's like, well, that's really good technology. Oh my God, it's getting easier. As the infrastructure as code becomes more programmable, it's becoming easier, more Lambda, more serverless action. You got new offerings. How are customers benefiting, for instance, from the three new offerings that you guys announced here? What specifically is the value proposition that you guys are putting out there? Yeah, so the big one is that, as we've tried to do with AWS over the years, customers get to focus on the things that really differentiate them and differentiate their businesses. So we take away, say in Redshift serverless, for example, all of the work that's needed to manage clusters, provision them, scale them, optimize them, and that's all been automated and made invisible to customers. So customers just think about data, what they want to do with it, what insights they can derive from it, and they know they're getting the most efficient infrastructure possible to make that a reality for them with high performance and low cost. So better results, more ability to focus on what differentiates their business and lower cost structure over time. Yeah, I had the Accenture guys on, they had brought this whole cloud continuum, which is their word for what Adam was saying, is clouds everywhere. And they're saying it's faster to match what you want to do with the outcomes, the capabilities and outcomes kind of merging together where it's easier to say, this is what we want to do and here's the outcome it supports. That's right. With that, what are some of the key trends on those outcomes that you see with the data analytics that's most popular right now and kind of where is that going? Yeah, I mean, I think what we've seen is that data's just becoming more and more critical on top of mind for customers and the pandemic has also accelerated that. We found that customers are really looking to data and analytics and machine learning to find new opportunities. How can they really expand their business, take advantage of what's happening? And then the other part is how can they find efficiencies? And so really everything that we're trying to do is we're trying to connect it to business outcomes for customers. How can you deepen your relationship with your customers? How can you create new customer experiences and how can you do that more efficiently with more agility and take advantage of the ability to be flexible in what is a very unpredictable world as we've seen? I noticed a lot of purpose-built discussion going on in the keynote with Swami as well. How are you creating this next layer of what I call purpose-built platform-like features? I mean, tools are great. You see a lot of tools in the data market. Tools are tools. If you're a hammer and you want to look for a nail, we've seen people overbuy too many tools and you have ultimately a platform. But there seems to be a new trend where this connect phenomenon was showing me that you got these platform capabilities that people can build on top of it because there's a huge ecosystem of data tools out there that you guys have as partners that want to snap together. So the trend is things are starting to snap together. Less primitive, roll your own, what you can do. But there's now more easier ways. Take me through that, explain the unpack that phenomenon of rolling your own primitives, which has been the way. Now too, here, here's some prefabricated software, go. Yeah. So it's a great observation, John, and you're absolutely right. I mean, I think there's some customers that want to roll their own and they'll start with instances, they'll install software, they'll write their own code, build their own bespoke systems and we provide what customers need to do that. But I think increasingly you're starting to see these higher level abstractions that take away all of that detail and mark as Adam put it and allow customers to compose these. And we think it's important when you do that to be modular, so customers don't have to have these big bang all or nothing approaches, you can pick what's appropriate. But you're never on a dead end, you can always evolve and scale as you need to. And then you want to bring these ideas of unified governance and cohesive interfaces across so that customers find it easy to adopt the next thing. And so you can start off, say, with batch analytics, you can expand into real time, you can bring in machine learning and predictive capabilities, you can add your natural language. And it's a big ecosystem of managed services as well as third parties and partners. And what's interesting, I want to get your thoughts while I got you here, because I think this is such an important trend and historic moment in time. Jerry Chen, who's one of the smartest VCs that we know from Greylock and Coincastles in the cloud, which kind of came out of a cube conversation here in the cube years ago, where we saw the movement of that someone's going to build real value on AWS, not just an app. And you see the rise of the snowflakes, the data bricks and other companies. And he was pointing out that you can get a very narrow wedge and get a position with these platforms, build on top of them and then build value. And I think that's the number one question people ask me. It's like, okay, how do I build value on top of these analytic packages? So if I'm a startup or I'm a big company, I also want to leverage these high level abstractions and build on top of it. How do you talk about that? How do you explain that? Because that's what people kind of want to know. It's like, okay, is it enabling me or do I have to fend for myself later? This kind of comes up a lot. Yeah, it's a great question. And if you saw Goldman's announcement this week, which is about building their cloud on top of AWS, it's a great example of using our capabilities in terms of infrastructure and analytics and machine learning to really allow them to take what's value added about Goldman and their position in the financial markets to build something value add and create a ton of value for Goldman by leveraging the things that we offer. To us, that's an ideal outcome because it's a win-win for us in Goldman, but it's also a win for Goldman and their customers. That's what we call the super cloud. That's the opportunity. So is there a lot of Goldman's opportunities out there? Is that just these unicorns? Are these like, I mean, that's Goldman Sachs. They're huge. Is this open to everybody? Absolutely, I mean, that's been one of the core ideas behind AWS was we wanted to give anybody, any developer access to the same technology that the world's largest corporations had. And that's what you have today, the things that Goldman uses to build that cloud are available to anybody. And you can start for a few pennies and scale up into the petabytes and beyond. When I was talking to Adam Slepsky when I met with him prior to re-invent, I noticed that he was definitely had an affinity towards the data. Obviously, he's Amazonian, but he spent some time at Tableau. So as he's running that company, so you see that kind of mindset of the data advantage. So I have to ask you, because this is something that I've been talking about for a while and I'm waiting for it to emerge, but I'm not sure it's going to happen yet, but what infrastructure is code was for DevOps and then DevSecOps? There's almost like a data ops developing where data as code or programmable data, if I can connect the dots of what Swami is saying, what you're doing is this is like a new horizontal layer of data, a freely available data with some government governance built in. That's right. So it's data is being baked into everything. So data is an ingredient, not a query to some database. It's got to be baked into the apps. That's data as code. That's right. So it's almost a data DevOps kind of vibe. Yeah, no, you're absolutely right. And you've seen it with things like MLOps and so on. It's all the special case of DevOps, but what you're really trying to do is to get programmatic and systematic about how you deal with data. And it's not just data that you have, it's also publicly available data sets and it's customers sharing with each other. So building this ecosystem of data and we've got things like our open data program where we've got publicly hosted data sets or things like the AWS data exchange where customers can actually monetize data. So it's not just data as code, but now data as a monetizable asset. So it's a really exciting time to be in the data business. Yeah, and I think it's so many tools. So I got to ask you, I got you here since you're an expert. Okay, here's my problem. I have a lot of data. I'm nervous about it. I want to secure it. So if I try to secure it, I'm not making it available. So I want to feed the machine learning. How do I create an architecture where I can make it freely available yet maintain the control and the comfort that this is going to be secure? What products do I buy? Yeah, so, you know, great place to start it as three. You know, it's one of the best places for data lakes for all the reasons that Swami talked about, durability, scale, cost. You can then use Lake Formation to really protect and govern that data so you can decide who's allowed to see it, what they're allowed to see, and you don't have to create multiple copies. So you can define that, you know, this group of partners can see A, B, and C. This group can see D, E, and F and the system enforces that and you have a central point of control where you can monitor what's happening and if you want to change your mind, you can do that instantly and all access can be locked down. You've got a variety of encryption capabilities with things like KMS and so you can really lock down your data but yet keep it open to the parties that you want and give them specifically the access that you want to give them. And then once you've done that, they're free to use that data according to the rules that you defined with the analytics tools that we offer to go drive value, create insight, and do something remarkable. That's lake formation and then you got Athena querying it so you got all kinds of tooling on top of it. That's right. You can have Athena querying your data in S3, lake formation protecting it, and then SageMaker's integrated with Athena so you can pull that data into SageMaker for machine learning, interrogate that data using natural language with things like QuickSightQ like we demoed. So just a ton of power without having to really think too deeply about developing expert skill sets in those tools. So the next question I want to ask you is because that first part of the great description. Thank you very much. Now 5G and the edge is here. Outpost, how is the analytics going on that as edge becomes more pervasive in the architecture? Yeah, it's going to be a key part of this ecosystem and it's really a continuum. So we find customers are collecting data at the edge. They might be making local ML or inference type decisions on edge devices or automobiles for example. But typically that data at some point will come back into the cloud, into S3 it will be used to do heavy duty training and then those models get pushed back out to the edge. And then some of the things that we've done in Athena, for example, with federated query, as long as you have a network path and you can understand what the data format or the database is, you can actually run a query on that data. So you can run real time queries on data wherever it lives, whether it's on an edge device, on an outpost, in a local zone, or in your cloud region and combine all of that together in one place. And I think having that data copies everywhere is a big deal. I got to ask you now that we're here at Reinvent. What's your take? We're back in person. Last year was all virtual, finally. It's not 60,000 people like a couple years ago. Still 27,000 people here. All lining up for the sessions, all having a great time. All good. What's the most important story from your area that people should pay attention to? What's the headline? What's the top news? What should people pay attention to? Yeah, so I think first off, it is awesome to be back in person. It's just so fun to see customers and to see, I mean, you, like we've been meeting here over the years and it's great that so much energy in person. It's been really nice. You know, I think from an analytics perspective, there's just been a ton of innovation. I think the core idea for us is we want to make it easy for customers to use the right tool for the right job to get insight from all of their data as cost-effectively as possible. And I think, you know, I think if customers walk away and think about it as being, it's now easier than ever for me to take advantage of everything that AWS has to offer to make sense of all the data that I'm generating and use it to drive business value. And I think we'll have done our jobs right. What's the coolest thing that you're seeing here? Is it the serverless innovation? Is it the new abstraction layer with data high-level services? In your mind, what's the coolest thing? And it's hard to pick the coolest. It's like a kick in the candy store. I mean, I think the, you know, the continued innovation in terms of performance and functionality in each of our services is a big deal. I think serverless is a game changer for customers. And then I think really the infusion of machine learning throughout all of these systems. So things like Redshift ML, Athena ML, QuickSight Q, just really enabling new experiences for customers in a way that's easier than it ever has been. And I think that's a big deal. And I'm really excited to see what customers do with it. Yeah, and I think the performance thing, to me, the coolest thing that I'm seeing is the Graviton 3 and the Graviton Progression with the custom stacks with all this ease of use is just going to be just a real performance advantage. And the costs are getting lower. So I think the EC2 instances around the compute is phenomenal. No, absolutely. I mean, I think the hardware and silicon innovation is huge. And it's not just performance. It's also the energy efficiency. It's a big deal for the future. We're at an inflection point where this modern applications are being built. And in my history, I'm old, my birthday's today, I'm in my 50s. So I remember back in the 80s, every major inflection point when there was a shift in how things were developed and mainstream client server, PC, internet work, you name it. Every time the apps change, the app owners, app developers all went to the best platform processing. And so I think that idea of system software applications being bundled together is a losing formula. I think you got to have that decoupling large scale. We're seeing that with cloud. And I think now, if I'm an app developer, whether I'm in a large ISV in your ecosystem or in the APN partner or startup, I'm going to go where my software runs the best, period. And where I can create value. I get distribution, I create value and it runs fast. I mean, that's what it wants. That doesn't cost me a lot. I mean, it's pretty simple. So I think the ecosystem's going to be a big action for the next couple of years. Yeah, absolutely right. And I mean, the ecosystem's huge. And I think, and we're also grateful to have all these partners here. It's a huge deal for us. And I think it really matters for customers. What's on your roadmap this year? What do you got going on? Can you share a little bit of trajectory without kind of breaking the rules of the Amazonian confidentiality? What's the focus for the year? What's next? Well, you know, as you know, we're always talking to customers and I think we're going to make things better, faster, cheaper, easier to use. And I think you've seen some of the things that we're doing with integration. Now you'll see more of that. And really the goal is how can customers get value as quickly as possible for as low cost as possible? That's how we went together in the long term. And yeah, we've always saved every time we see each other. Data is at the center of the value proposition. I've been saying that for 10 years. Now it's actually the value proposition powering AI and you're seeing because of it, the rise of super clouds. And then these super clouds are emerging. I think you guys are the underpinnings of these emerging super clouds. And so it's a huge trend. I think the Goldman Sachs things are validation. So again, more data, the better. That's right. Cool things happening. It is just, it's everywhere. And the diversity of use cases is amazing. I mean, I think from the Australia Swimming Team to Formula One to NASDAQ, it's just incredible to see what our customers are doing. Can't wait to see the great route. Great to see you. Thanks for coming on theCUBE. Thanks for being here. Pleasure to be here as always, John. Great to see you. Thank you. Thanks for sharing. All the data is the key to success. Data is the value proposition. You're seeing the rise of super clouds because of the data advantage. If you can expose it, protect it, govern it, unleash its creativity and opportunities for entrepreneurs and businesses, of course. You got to have the scale and the price performance. That's what Amazon is doing. It's theCUBE coverage. You're watching the leader in worldwide tech coverage here in person for AWS re-invent 2021. I'm John Furrier. Thanks for watching.