 Live from Las Vegas, it's theCUBE! Covering AWS re-invent 2019. Brought to you by Amazon Web Services and Intel, along with its ecosystem partners. Well, welcome back everyone. It's CUBE's live coverage in Las Vegas for AWS re-invent. It's theCUBE's 10th year of operations, our seventh AWS re-invent. And every year it gets better and better. And every year we've had theCUBE at re-invent. Jerry Chen has been on as a guest. He's a VIP Jerry Chen. Now a general partner at Greylock Tier One, one of the leading global venture capitals at Silicon Valley. Jerry, you've been with the journey with us the whole time. I guess I'm your good luck charm. Well, keep it going. Keep on changing the game. So thanks for coming on. Thanks for having me. So now that you're a seasoned partner now at Greylock, you got a lot of investments under your belt. How's it going? It's great. I mean, look, every single year I look around the landscape thinking what else could be coming? What's going to be surprising this year? What's the new trends? What both macro trends, also company trends? Like who's going to buy who? Who's going to go public? Every year just gets busier and busier and bigger and bigger. All these new categories are emerging with this new architecture. I call it cloud 2.0, maybe next-gen cloud, whatever you want to call it. It's clear visibility now into the fact that DevOps is working. Sure. Cloud operations, large-scale operations with cloud is certainly a great value proposition. You're seeing now multiple databases. Pick the tool. I think Jassy got that right in his keynote. I believe that. But now the data equation comes over the top. So you got DevOps, infrastructure as code. You got data now, looking like it's going to go down that same path of data as code where developers don't have to deal with all the different nuances of how data is stored, how it's handled, where is it? Warm or cold or at glacier? So developers still don't have that yet today. This seems to be an area of Amazon. What's your take on all this? I think you saw, so what drove DevOps? Speed, right? It's basically how developers should also do operations merging into groups. So we're seeing the same trend, data ops, right? How data engineers and data scientists can now have the same speed developers had for the past 10 years, data ops. So A, what does that mean? Give me the menu of what I want, like Goldilocks, too big, too small, just right. Too hot, too cold, just right. Like give me the storage tier, the data tier, the size you want, the temperature you want, and the speed I want. So you're seeing data ops get the same kind of Goldilocks treatment as developers. And on terms of like cloud evolution, again, you've seen the movie from the beginning at VMware now through Amazon seventh year. What jumps out at you? What do you look at as squinting through the trend lines and the fashion of the features? It still seems to be the same old game. Compute networking storage and software. Well, I mean, compute memory storage, those are the atomic building blocks of compute, right? So regardless of serverless or these high-level frameworks, deep down, you still have compute networking and storage. So that's the building blocks. But I think we're seeing 10th year re-event, this kind of, it's not one size fits all, but this really big, fat, long tail, small instances, micro instances, serverless, big instances for like jumbo VMs, bare metal, right? So you're seeing not one architecture, but folks can kind of pick and choose, buy compute by the drip, the drop, or buy compute by the whole VM or whole server fold. And a lot of people are like the builders love that. Amazon owns the builder market. I mean, if anyone who's doing a startup, they pretty much start on Amazon. It's the most robust, you pick your tools, you build. But Steve Mulaney was just on before, it says enterprise don't want power tools, they're going to cut their hand off, right? So Microsoft's been winning with this approach of consumable cloud. And it's a nice card to play because they're not yet there with capabilities with Amazon. So it's a good call, they got enterprise sales force. Microsoft playing a different game than AWS, because they have to. Sure, I mean, what's the football analogy? You have no running game, you need a passing game, right? So if you can't beat them with a running game, you go with a passing game. And so Amazon has kind of like the fundamental building blocks of power tools for the builders. There's a large segment of population out there that don't want that level of building blocks, but they want it a little more prescriptive. So I think Microsoft has been around enterprise for many, many years. They understand prescriptive tools and architectures, so they're going to come a little bit more prefab, if you will. Here's how you have to construct your right application, ML apps, AI apps, et cetera. Let me give you the building blocks at a higher level of abstraction. So I want to get your take on value creation. Sure. If it's still early innings, it's still got a lot more growth. You start to see Jassy even admit that in his keynote today. He said, quote, there are two types of developers and customers, people who want the building blocks and people who want solutions or prefab or some sort of more consumable. It's a little prescriptive, yeah. So I think Amazon's going to start going that way. But that being said, there's still opportunities for startups. You're an investor, you invest in startups. Where do you see opportunity? If you're looking at the startup landscape, what is the playbook? How should you advise startups? Because, you know, of course, I have the best team whenever, but you look at Amazon, it's like, okay, they got large scale. Yeah. I'm going to be a little nervous. So they going to eat my lunch. Do I take advantage of them? Do I draft off them? There are white spaces as vertical markets exploding. Now available. What's your view on how startups should attack the wealth creation opportunity value creation? They're, I mean, Amazon's creating a new market, right? So you look at their list of many of services. There's just 175 services out there, which is basically too many for any one company to win every single service. So, but you look at that many of services, each one of the services themselves can be startup, our collection of services can be startup. So I look at that as a roadmap for opportunity of companies can actually go in and create value around AI, around data, around security, around observability. Because Amazon is not going to naturally win all those markets. What they do have is distribution, right? They have a lot of developer mind share. So if you're startup, you play one or three themes. Like one is how do I pick one area and go deeper IP, right? I cheaper, better, faster, own some IP and they're going to execute better. And that's doable over, over again in different markets. Number two is, we've talked about this before, it's not going to be one cloud wins all. Amazon's clearly lead, they have won most of the cloud so far, but it'll be a multi-cloud world. It'll be on-premise world. So how do I play a multi-cloud world? It's another angle. So go deep in IP, go multi-cloud. Number three is this end-to-end solution, kind of prescriptive. I think Amazon will get you 80% of the way there, 70% of the way there, but if you're like an AI developer, you're a CMO, you're a marketing developer, you kind of want this end-to-end solution. So how can I put together a full suite of tools from beginning to end that can give me a product that's a better experience? So either I have something that's deeper IP, play a scene between multiple clouds or give an end-to-end solution around a problem and solve that one problem for a customer. In most cases, on the underlays Amazon or Azure. Or Google or Ali Cloud or on-premises, not going away anytime soon, right? And so how do we create a single fabric, if you will, that looks similar? I want to riff with you in real time here on theCUBE around data. So data scale is obviously a big discussion. That's starting to happen now. Data tsunami, we've heard that for years. So there's two scale benefits. Horizontal scale of data and then vertical specialism, vertical scale, or using AI machine learning with in-apps, having data. So how do you view that? What's your reaction to the notion of creating horizontal scale value and vertical specialism value? Both are a great place for startups, right? They're not meet exclusive. So I think if you go horizontal, the amount of data being created by your applications, your infrastructure, your sensors, time series data, a ridiculously large amount, right? And that's not going away anytime soon. I recently did an investment in Chronosphere that you guys covered over at KubeCon a few weeks ago that's talking about metrics and observability data, time series data. So they're going to handle that horizontal amount of data, petabytes and petabytes. How can query this quickly, deeply with a lot of insight? That's one play, right? Cheaper, better, faster at scale. The next play, like I said, is vertical. It's how do I own data, slice the data with more context than no one else can have? We talked about the virtual cycle of data, right? The system of intelligence, if you will. If I own a set of data, be healthcare, government, or self-driving car data that no one else has, I can build a solution end to end and go deep. And so either pick a lane or pick a geography, you can go either way. It's hard to do both, though. It's hard, for startup. For startup. Even any big company. Very few companies can do two things well. Startups especially, it's a seed by doing one thing very well. I think my observation is that I think in looking at Amazon is that they want the horizontal and they're leaving up on the table for our startups the vertical. Yeah, if you look at their strategy, the lower level Amazon gets, the more open source, the more ubiquitous you try to be from containers, serverless, networking, S3, basic substrate, so horizontal, horizontal, low price. As a good hire up from deep mind, like AI technologies, perception, prediction, they're getting a little more specialized, right? So you can see these solutions around retail, healthcare, voice, so the higher up in the stack, they can build more narrow solutions because like any startup, any product, you need the right wedge. So what's the right wedge in the customers? At the base layer for the developers, building blocks, ubiquitous. For solutions, marketing, healthcare, financial services, retail, how do I find a fine point wedge? So the old venture business was all enamored with consumer over the years. And then what, maybe four years ago, enterprise got hot, we were lowly enterprise guys, when no one- Enterprise been hot forever, my mind, Tom, but maybe- Well, first of all, we've been hot on enterprise, we love enterprise. But then all of a sudden it just seemed like, oh my God, people had an awakening, like and in this real value that we had, the IT spend has been trillions, and the stats are roughly 20 or so percent. Yet to move to the cloud or this new next-gen architecture that you're investing companies in. So a big market, that's an investment thesis. So a huge enterprise market, Steve Mulaney from AVH has called it a thousand foot wave. So there's going to be a massive enterprise money, big bag of money on the table. Lot of retransformations, a lot of reborn in the cloud, a lot of action. What's your take on that? Do you see it the same way? Because they're getting in big time, Goldman Sachs on stage here. So that's a lot of cash. How do you think it's going to be deployed and who's going to be fighting for it? Well, I think, we talked about this in the past, when you look to make an investment, you want to pick as a startup founder or as a VC, you want to pick away bigger than you, bigger than your competitors, right? So on the consumer side, the classic example, you're Instagram fighting, Facebook and photo sharing, you pick the mobile first wave, iPhone wave, right? The first mobile native photo sharing. If you're fighting enterprise infrastructure, you pick the cloud native wave, right? You pick the big data wave, you pick the AI wave. So first as a founder startup, I'm looking for these macro waves that I see not going away time soon. So moving from batch data to streaming real-time data, that's a way that's happening, that's inevitable. Dollars are flowing from slower batch databases to streaming real-time analytics. So Rockset, one of the investments we talked about, they're writing that way from going batch to real-time, how to do analytics and SQL on real-time data. Likewise, time series is going from like, batch data, slow data to mass amounts of time series data, chronosphere, playing that way. So I think you have to look for these macro waves of cloud, which we, anyone knows, but then you pick these small wavelets, if that's a word, like a wavelets or a smaller wave within the wave that says, okay, I'm going to pick this one trend, right as a startup, right as investor, and because that's going to be more powerful than my competitors. And then get inside the wave or inside the tornado, whatever metaphor. We're going to torch the metaphors, but yeah, ride that wave. All right, Jerry, great to have you on, seven years of CUBE action, great to have you on. Congratulations, you've been with us the whole time. Congratulations to you, the CUBE, the entire staff here. It's amazing to watch your business grow the past seven years as well. And we soft launch our CUBE 365 search hits in Amazon's marketplace. SAS, our first SAS offering. I love it, I mean. No venture funding, you know, and we're going to be out there, you know, maybe we'll let you in on the deal. But now, like you broadcast the deals with the rest of the market. Jerry, great to have you on, again, great to watch your career at Greylock, always having on, great commentary, awesome time. Jerry Chen, venture partner, general partner at Greylock, it's a CUBE coverage, breaking down the commentary, extracting the signal from the noise here at ReInvent 2019, I'm John Furrier, back with more after this short break.