 Welcome to theCUBE, CUBE Conversations here in Palo Alto. I'm John Furrier in the studio is Sri Ramathon, who's the group vice president mobile cloud with Oracle. Here to talk about the landscape around mobile AI and software, how it works end to end and really the realities of it. Welcome to theCUBE conversation. I'm glad to be here. So we just talked before we went on camera about chatbots and I was, I've been very critical of chatbots. I mean, certainly who doesn't like a chatbot that's got a little glam to it. But they've been seen gimmicky. Okay, you guys been doing a ton of work with chatbots at developer week and in this code program you're running. Chatbots is a nice face, but to make them really work, it takes time. Do you have thoughts on that in general? Absolutely, absolutely. The digital landscape is going through a very interesting set of transformations and people would traditionally use their browser to access business applications or consumer applications. That shifted to mobile and frankly, there's just too many apps. There's thousands of apps and if I have a daughter, she's a mall rat, there's eight stores in the mall that she likes. She's not going to download those eight apps. It's not going to happen. She lives her life in the context of four apps, Snapchat, Instagram, Facebook Messenger, and maybe WhatsApp and she communicates and collaborates and entertains herself through the context of those apps and that's where I think a smart bot would make a huge amount of sense and I think the difference to making a bot successful is one that it has to solve your problem, whatever it is quickly. It has to be smart, be predictive and try to predict what things you need and be right most of the time. And from a developer perspective, you have to have the tools and the runtimes to provide the right experience and I think these are all kind of the reasons why this first generation of chatbots about which you've been critical and other people have been critical, they've been justified in their criticism because they haven't met these basic bars. They're not kind of easy to build and deploy with the right UX. They're not smart and they're not easy to use. Those are all like big problems. So I think those are the problems we solve with our bot platform. And AI is certainly very early street but before they know that's a good comment. I want to drill down on some of the back end, what unifies and what will make a chatbot like experience scale. But before we do that just take a minute to explain what you do at Oracle because I think you're an interesting position. You're on the engineering side on the mobile portfolio. What is that, what does that mean? Absolutely, so there's kind of three things we do. We build an embass, a mobile backend as a service and this mobile backend as a service gives you a set of tools and runtimes that make it easy for you to build, deploy and run and manage mobile applications. We give you technology that makes it very easy to find back end services and build mobile optimized APIs. We build technology that makes it very easy for you to build apps that work offline, that store data offline and do full duplex synchronization which is very important for a lot of enterprise scenarios. We build a push gateway that lets you send notifications and content to applications very fast and rapidly. And then the second thing we do is we have some analytics infrastructure that give you deep insights into how, where, when and why a mobile app is being used. And so if you're a marketer you want to know what features of the app are being used and what are not being used and how they're being used and how to improve your apps. So there's a DevOps workflow associated with mobile analytics and kind of the third thing we do is we build this bot's platform that lets you build a cross-platform bot that you can then deploy on Alexa, SiriKit, Facebook Messenger, Slack or any channel that you want. So once you define a bot, you know, you then are able to deploy it pretty much to anywhere you want and it uses the same APIs, the same API and integration stack in the back end. So you're now able to expose your enterprise data and services not just to web apps and to mobile apps but also to bots. So it's an integrated stack that lets you, you know, design, build, run and manage web apps, mobile apps as well as bots, all with the same consistent set of tools, front times and management. You know, the app fatigue is one of the things that comes out of that. When I hear you talk like that, I can see that connecting the dots to app download fatigue. I mean, I just don't like downloading apps anymore but I have to as part of a service I need to use. I mean, the classic example is on the entertainment side, sports, NCAA March Madness, you use that all the time and then it's gone, I never use it again. And then it goes away, NFL's over, I was using that for a while, there's no more sports or checking my airline reservation but I now fly multiple airlines. It's like, what the heck is going on? It's a crappy user experience. You are making the case for a bot. I mean, you just made it very well. The thing though is that, you know- Those are different companies. So now I need APIs, I need integration, data governance, all this is a bunch of details that is not a Greenfield consumer app. So how do we make it all work? What's a lot of work? See what's happening is, in the old world, you had all these clients of applications and the browser became the portal where all these apps, it became kind of this meta app where you just got everything. All of your apps came through that. The messaging platforms are becoming that. They're taking all of these mobile apps that exist and they're aggregating their most important capabilities to start with, not all their capabilities, their most important and germane capabilities and they're now being surfaced as bots within the context of the messaging app. And what's happening is that the experience today that you see in a bot, which is one of the criticisms I think you are alluding to is simply not there. You don't get the NFL mobile experience that you get on the NFL app through an NFL mobile bot maybe. And that's where a good amount of work is going in. How I can elevate the experience, make it fantastic, but at the same time take all the capabilities that are in all of these apps and surfacing them through the bot channels, whether it's a voice channel or a messaging channel. So the bots are essentially just an indication of the evolution of where the software is. So now hardening that and making that kind of horizontally scalable because that at the end of the day, from my view, and I always said this is something that I think you guys are doing right at Oracle and I don't think anyone's picked up on this yet is that you guys have been the master at the vertical stack Oracle has been as traditional company, that's the way it was. Everyone had their own stovepipes or own silent stacks, but the cloud now is a horizontally scalable architecture. And so now you have to have, but that also dilutes the specialization involved in software, which is app specific. So what's interesting is there's a new dynamic between horizontally scalable, but yet vertically integrated. What's your thoughts on that? Because mobile really teases it out. All the action right now is at the top of the stack. Is that where the vertical specialization is? And then this is the chat bot layer, are going to be the traversal? Is it going to be the boss? I mean, where does this all fit in your thoughts? That's a great point. Look, I think you're always going to have apps, which is part of the SaaS portfolio that Oracle has, that have deep business knowledge and ontology and very specific capabilities, data models that are unique to their domain. And our PaaS layer, which is the team that I belong to, you know, is a set of capabilities that all those apps can use first, or that anybody who wants to extend and customize those apps can then leverage. So a lot of those common PaaS capabilities fall into that. And mobile and bot are an example of that, because each one of these underlying capabilities, which is where the value is, a lot of these SaaS apps, they need the best possible digital channel to get their surface, their capabilities out, right? Or to take existing capabilities that they may ship with as part of their out-of-the-box experience and enhance and extend those. Sree Ramathan, who's here inside the CUBE conversation, I'm going to ask you the tough question that's on everyone's mind, and it's on my mind too as well. But I think, you know, most developers that aren't in the Oracle family yet are out in the wild, they're out in the cloud native, they're doing some cool things and they're developing apps. And you know, the classic developer psychology is, I want to work on really cool things and I want it to be relevant and I want it to be successful. And success is defined by either distribution on open source or some sort of monetization capability with traffic and cash. The question is, how does Oracle help that? Because you know, I see a huge opportunity for you guys where you guys, and if you look at Microsoft right now, they're essentially satisfying their own stuff and you are too, but the money-making capability that you guys could provide to a developer through your existing business is phenomenal opportunity. So as a developer, I would think, hey, I would love to partner with a company that's got great client base. And so how do you have that conversation? Or I mean, developers don't like to talk money, it's not about greed, but in the day it's the elephant in the room, you know, how do I be successful with Oracle? Whether it's financial money-making or traction and open source or that relevance and the distribution of the product. That's a great question. Look, I think from our perspective, our goal, our fundamental goal when it comes to developers is to make our software dead simple to use. Right? It's to embrace open source, right, which we've done very successfully. So you can use any programming language that you want, right? We give you the best infrastructure to manage your development environment, right? Your development life cycle, so to speak. Whether it's basic things like, you know, how we manage your code on your instance of a Git, you know, how we, you know, provide hooks into all of the common CICD infrastructures out there, you know, how we give you application performance management tools after you've built your app and how we give you the best container cloud service so that you can very easily deploy your app and get it up and running, you know, really fast and not have to worry about things like security and scalability and, you know, all those things are fundamental features of the cloud fabric. That even me as a service owner in Oracle, we rely on those same underlying capabilities at the road that the superhighway underneath us provides. So I think the way to get to developers is to give them technology that makes it very easy to build, you know, run and manage their applications. And I think that's what we're focused on. If you had to answer the question, if I asked you, okay, Shree, tell me the building blocks, okay? I like to think in terms of building blocks, what's the fundamental foundational core element of the mobile as a service that you guys are providing? Well, our runtimes are all fundamentally based on web logic. What we expose to developers are all, you know, based on Node.js. So if you are a mobile developer and you're building mobile apps on our infrastructure, the programming language you'd use is actually Node. It's not even Java. It's pure Node, you know, and we give them a browser-based interface that lets them use any editor they want to build all the custom code that they require in JavaScript and upload it straight to our Node container. Our Node container is part of our application container cloud service. So it's an ACCS model where you get to, you know, manage your containers. You can set up autoscale seamlessly and you get all of those capabilities right out of the box. And all of our capabilities, you know, it's not all of them. So if I have an app and I have a flash mob coming in, you guys on your cloud are doing all the autoscaling. Absolutely. You can set it up so that... Queuing and all that stuff built into it as well? It's all built right into it and you'll be able to leverage that off the bat and not have to worry about it. So I don't need to worry about having five zillion IT guys on provisioning any kind of hardware of any kind. Autoscale's on its own. Exactly. Can I set parameters around that? Is that how it would work? You can select the scale characteristics you desire and then we also let you deploy across geographies. You know, we just announced our data center expansion. You can deploy it right across geographies. You can identify availability zones and disaster recovery zones so that you can make sure that your data and your customer are always safe. Okay, you got AI around the corner, which to me is a great little mental model for folks to kind of understand IoT and mobile because IoT is Internet of Things and certainly Tesla's are driving software and everyone can relate to the physical world but no one can really grok IoT. What is IoT? Air conditioning? It's like, it's kind of boring, kind of feels boring but it's pretty relevant. AI gives a nice picture of that. Oh, automated intelligent machines talking, flying saucers, flying cars. It can get a nice mental model. So with that in mind, I got to ask you, what is the coolest thing you're working on right now that fits into that sci-fi futuristic technical coolness? Can you build a bot on our platform? You get access to your own slice of a neural network and it mimics the way a human brain works and the thing about it is you don't have to understand all of the deep learning algorithms were like implementing, right? You just use it, right? We make it dead simple for a developer to model a bot so the bot can understand intent very simply, right? You provide it with some training data that's very simple. We make it so simple in fact that you can be up and running with a bot that's live in a few hours. That's how simple it is. So I think the coolest thing is to be able to work on a practical application of AI. Like you said, AI is this buzzword that's thrown around but we're actually giving you a slice of your own neural network in a very easy to consume way and I think that's what I'm most excited about. I think one of the things that comes out a lot and the word adaptive computing has been thrown around all these cliches are out there but what you're kind of getting at in this new world is essentially algorithms have been out there so algorithms can be determined based upon the data available so ideally you don't really, you should never be in the algorithm coding world that should be algorithms can apply to situations. Is that kind of what you're getting at here? Oh absolutely, I think look, I don't think we're in search of a problem to solve. I think we're in search of like practical solutions that we can apply. We're not in this AI business for the sake of being in it. We've got a very practical set of use cases that will tactically add value to our customers. Yeah, device data and then data cloud, you guys have both that going on and that will be a good input to the AI. Absolutely. Ashree, thanks for coming in and sharing the vision and the commentary around your efforts, appreciate it and sharing the color here on theCUBE. I'm John Furrier, you're watching CUBE Conversations here in Palo Alto. Thanks for watching.