 Ladies and gentlemen, please welcome Senior Vice President of Amazon Web Services, Andy Jassy. Welcome to our third annual AWS re-event. It is awesome to be here, and we really appreciate your spending a few days with us. This is our favorite time of the year in AWS when we get to spend these days with you. We have about 13,500 attendees here from 63 countries around the world, and another 15,000 listening in on the live stream. Over the next few days, we're going to teach about 275 breakout sessions in classes and boot camps, and the reason we do this is that this conference is different from most of the technology conferences. This is not a sales and marketing conference. This is an education and learning conference, and our hope is that you learn a lot from each other. Those 275 classes, two-thirds of those classes will either be taught by customers or partners or feature them, and so what we hope is you learn more about the cloud, you learn from each other, and then you take that learning back to your own companies and change your businesses and the customer experiences in your businesses. So we're very excited you're here. Thank you for being here. We're going to try and get right to it. We have a lot to share with you, so giddy up. I thought I would start by sharing a quick update on how the AWS business is progressing and the business is growing really dramatically, and I'll give you just a few data points. At this point, we have over a million active customers in AWS, and we define an active customer as a non-Amazon customer who's using the platform during a month. And a couple additional data points of scale are storage service, which is Amazon S3. You can see this tracks the year-over-year amount of petabytes transferred to and from S3. And this is a pretty good way to look at the usage and the size of the service because it not only tracks the number of objects but then the real activity going on with the service. And you can see on an already large base that S3 has grown 137 percent year-over-year. And the same is true in our compute service, EC2. This tracks the number of instance hours run in EC2 year-over-year, and it's growing 99 percent already on a very large base, so the business is growing very quickly. We have at this point, as I mentioned, over a million customers using AWS in every imaginable business segment and every imaginable company size, and it ranges from startups like some of the most successful startups like Pinterest and Dropbox and Airbnb and Supercell and Spotify and the most recent startup to move to AWS Secret to large enterprises all over the world. So in oil and gas, Shell and BP and HES and financial services, FINRA and Intuit and SunCorp and the Commonwealth Bank of Australia and HealthCare, J&J and Pfizer and Bristol Meyers, Squibb and Merck, and the communication space, Vodafone and Comcast and NTT-Docomo, and the technology arena, Netflix and Samsung and Adobe, and the media space, Condi Nast, News Corp, Dow Jones and the New York Times and the Weather Channel. Every imaginable business segment is using AWS in a very meaningful way. And then we also have built a pretty large public sector business where we have over 900 government agencies all over the world using AWS. These are agencies like the Navy and NASA and JPL and the intelligence community. And then over 3400 academic institutions that are also using AWS. So a very broad and diverse customer base that continues to grow very quickly. And then we have a very large and vibrant ecosystem of partners that are helping our customers move to the cloud and use the software they're used to using just on top of AWS's technology infrastructure platform. So we have thousands of systems integrators that have built practices on top of AWS. And these range from very large ones like Accenture and Cognizant and Capgemini and Booz Allen and Infosys and Wipro to SIs that were largely born in the cloud that are doing a lot of the work right now for enterprises. These are SIs like Second Watch and CloudNexa and Slalom and Bulletproof and DataLose. And then we also have thousands of ISVs that have allowed their software to run well on top of the AWS technology platform. And again, a very large number here. They range from companies like Acquia and Infor and Informatica and Heroku and Adobe and Autodesk and NetApp and SAP and Splunk, a very large and vibrant ecosystem to help you move to the cloud. And then about two years ago, we launched an AWS marketplace. And this is a place where ISVs can put their products in a tightly integrated way in front of AWS customers, and AWS customers can find products from our partners that allow them to deploy easily. And at this point, we have about 1,900 products across 23 categories. And just to give you an idea of the scale of this, just last month alone, our customers ran 70 million hours of products from this marketplace just in the last month. This is an area of our business that's growing very dramatically as well. This is an interesting slide. We looked at the large enterprise technology companies in the world and looked at their relative revenue growth rate year over year. And what we found was that there were some companies, a couple of them that were growing in the 20-ish percent year over year range. Most were growing in the single digits. Some like IBM are shrinking. But if you look at AWS at well over 40% growth year over year, it is the largest multi-billion enterprise IT company in the world. And it's why when you look at relative comparisons of the players in the technology infrastructure space, AWS ends up where it does in a lot of these analyses. This is the most recent infrastructure as a service, Magic Quadrant that Gardner did. And you can see AWS in the top right. And you can see the other 14 companies clustered somewhere toward the middle there. And Gardner's assessment was that AWS, the overwhelming market share leader with more than five times the compute capacity in use, than the aggregate total of the other 14 providers combined. So it's a pretty significant leadership position that's continuing to expand. And what you've seen over the last few years, but particularly accelerating in the last 12 months, is that more and more companies are moving their workloads in large amounts to AWS into the cloud. This is a quote from the distinguished analyst at Gardner Lydia Long. And she said, quote, fascinatingly, my Orlando Symposium 101 seemed to be heavily slanted towards all in migrations to public cloud, IAAS, mostly AWS. But here's another one from Crawford Del Prat from IDC, and he said, by 2020, the distinction between public and private cloud disappears as self-built private clouds become extinct. Now, you can have different opinions about how complete and how fast this transition to the cloud is going to happen. But it also seems pretty apparent at this point that cloud has become the new normal, that companies of every size are deploying new applications by default to the cloud. And then large companies are trying to figure out how to migrate as many of their applications as fast as possible to the cloud. In the old world, most companies had their own infrastructure on premises, and they had very little functionality, a little bit of compute, a little bit of storage. All the features that you wanted to add to these services, teams had to build themselves. It was hard to get any change accomplished, and it was really slow. And so what you had were these companies spending millions of dollars for expensive and flexible, slow-living infrastructure that's kind of frozen in time. The new normal is that companies have this really large, really robust, fully-featured technology infrastructure platform from AWS at their fingertips so they can get from idea to launch as fast as possible. And this is a very broad and expansive platform. We have regions in 11 different regions all over the world, and a region for us is a place where we have multiple data centers. And that's a little bit different from what a lot of the other providers do. They'll have a single data center and call that a region. Our customers want to run their applications in a fault-tolerant way, and so that means they want to deploy their applications across multiple data centers. So all of our regions have multiple data centers. We have 28 availability zones across those 11 regions. And then the platform has all kinds of functionality. It's got compute, it's got load balancing, it's got auto-scaling, it's got multiple types of storage, object storage and block storage and archival and backup storage, it's got a content distribution network, it's got a relational database service, it has a NoSQL database service, it has a caching capability, and then all kinds of networking to allow you to optimize your security and performance on the platform. And then we have identity federation, identity and access management. So you have fine-grained access control, all kinds of ways to audit your usage on the platform, and then monitoring and logging for you. And then all kinds of platform services. So in the analytics space, if you wanted to do a dupe on the platform, you can if you want to do processing of streaming data in real time, you can do that. If you want to use our data warehouse, have at it. We have pipelines that allow you to safely move data inside of our platform to different services. We have a queuing service, we have notifications, we have email, we have transcoding, we have search, we have app streaming. We have a number of mobile capabilities for you. We have an identity database and sync and analytics. And then a mobile push service, over 10 billion push notifications are being sent a week on top of AWS at this point. We have enterprise applications. We've launched a virtual desktop in the cloud last year at ReInvent called Workspaces. And we've launched an enterprise storage and collaboration service this summer. And then a number of people services that you have told us you wanted us to build the last few years to make it easier to move to the cloud. And these range from not only account management but from premium support where you get a one-on-one technical account manager that knows your footprint on AWS and helps you move. We have solutions architects and professional services. We have training and certification. Very, very broad and expansive technology infrastructure platform. But it's not only having a lot of services that distinguishes the platform. It's having a lot of features within those services. So I'll give you a few examples. Sometimes people get a little confused and they say, well, so-and-so has a compute service. Isn't that the same as any other compute service? Well, if you look at compute and you look at all the different workloads that our customers want to run in our platform, it requires different ratios of the components inside a server to allow them to optimally run those workloads. And so we started with a general purpose compute family that had a comparable ratio of memory and CPU. And then we added a CPU optimized family and a memory optimized family and a storage optimized family and an IO optimized family and a GPU optimized family. And then our latest family, we launched in the summer, which is our T2 family, which is our lowest cost burstable performance family. Nobody else has half of these instance families. Or you look at storage. It's never been easier and less expensive to store data. And a lot of our customers wanted to be able to retain data longer than they could afford to do so in the past. So we built an archival and backup storage service called Amazon Glacier that's a penny per gigabyte per month. Totally changes what kind of data you can store. Or our block store, EBS, we started with a magnetic offering. And then we had a lot of customers who had throughput demanding workloads that wanted more consistent and a higher amount of IOPS. So we've offered a provisioned IOPS or PIOPS offering that allows customers to provision how many IOPS they want. And then we deliver 99.9% of the time to that provision IOPS number. That's what we guarantee customers. And then we just launched over the summer a new general purpose volume type in EBS, which is now the default, which is SSD back and allows you to burst with channels virtually any type of workload you need. If you look in the storage space, nobody has an archival and backup storage service or the options that we have in EBS. Or look in the database space. We have a relational database service. And we have it in four different database engines. We have MySQL, Oracle, SQL Server, and Postgres. There's nobody who has more than one of these engines. Or look in the data warehouse space. Apart from being the only technology provider with a data warehouse solution for the cloud, this team has delivered 95 additional features and capabilities to Amazon Redshift, our data warehouse service, and the 20 months since we've launched. So it's a very broad amount of services and then a lot more depth of features within those services than you'll find anywhere else. And probably the thing that the team is collectively most proud of besides the security and the operational performance of the platform in the first eight and a half years is the pace that we've been able to continue innovating on your behalf. So just for perspective, in 2012, we launched what we considered 159 significant services and features. In 2013, that number was 280. As of Monday, that number is 442 year to date. And we think we're on our way to a little bit over 500. So the pace of innovation is accelerating in AWS. And whether it's a data warehouse solution that's very high performance at one-tenth of the cost of the traditional data warehouse providers, or a very fast, high throughput, low latency, no SQL database in DynamoDB that's becoming the de facto standard for gaming and mobile apps, or a brand new platform for processing streaming data in real time in Amazon Kinesis, the breadth of the platform and the innovation of the platform is allowing customers to do things that they'd really only dreamed about and talked about but never found viable to do. And a good example of that is the first customer that I'm going to bring up. And this is a customer we've been working with for several months. It's Major League Baseball. And Major League Baseball, I hope some of you are baseball fans. I'm a big baseball fan. And they're pretty exciting last few weeks, too, with the World Series that culminated with the Game Seven, the Giants knocking off the feisty Kansas City Royals in Game Seven. But the thing that people, oh, there's some giant fans. The thing that people don't realize sometimes unless you watch carefully is that Major League Baseball is one of the most sophisticated technology companies in the world. Their streaming capabilities are better than almost anybody anywhere. And they do more to improve their customer experience and their viewers' experience than not only other sports, but really more than almost any other enterprise. It's pretty amazing. So to talk to you about a brand new application that they've built that changes the way you're going to watch games, it is my pleasure to bring up the executive vice president of CTO of Major League Baseball, Joe Insarilla. Of the middle, Tejana makes another, makes an incredible catch. I'm the CTO for Major League Baseball fans' media. As Andy said, there's a lot of people here from other countries, different parts of the world. Hopefully most of you recognize the silhouetted batter. But for those who don't, Baseball's like cricket, but faster. I threw the advanced media out there. So what's Major League Baseball advanced media? And I'd like to just give you a couple of brief history of us. And so back in 2000, the commissioner and the owners got together and they said, look, if Baseball's going to innovate in technology, we're not going to do it inside a 125-year-old institution. So what we need to do is create a technology company whose mission is to bring Baseball into the digital age. And that's what they did. So starting from scratch in 2000, with a little bit of cash and some rights, we have built a business that's now worth about $6 billion. Most of that is on our consumer product, like at Bat, which you can get on Amazon Firephone and all that sort of fun stuff, as well as our streaming product, like MLB.tv. But we also do things for other folks like ESPN and CBS, Turner, other people in the marketplace. And we've turned our platform into a service, as well. All of this has a commonality of what's happening. Consumer behavior is changing. It's going more online. It's going more mobile. Preferences of customers. A lot of people would rather watch on their iPad than on a big screen TV. And we want to be there. We absolutely think that it's crucial to the game to continue to evolve and serve the fan where they want to be served, when they want to be served. And all of that is putting increasing demands on IT, compute, the proliferation of devices, data, all sorts of stuff that's really driving innovation in our industry. And so one of the things that we're working on that I think is the most exciting, cool thing I've worked on in my career in baseball is Stackcast that's actually powered by AWS. And quite simply, what we're trying to do is, for the first time, measure things we've never been able to measure before. So baseball has some great statistics about offense. We have some great statistics about pitching. We don't have all that many great statistics about defense, and that's simply because the game is complex and hard and people move and the ball goes really fast. And so how are we going to address that? How can we describe these things in an empirical way? So we can say that this guy made a play that this guy would not have made a play and be able to do that with some amount of scientific prediction and accuracy. And so that's what we want to do. The question is, how are we going to get there? And what technology are we going to use to get us into that new era? So we'll take a look here at a little simulation we have that shows the way that the technology works. Essentially, what we decided through a lot of testing was that we weren't going to be able to just capture this data with one thing. And unlike a lot of other leagues, we've decided that we didn't want to change the game, no augmentation into the ball, no augmentation of the players. And so we start with a radar tracking system that was actually designed for missiles that can sample the ball 2,000 times a second. Down to the resolution, we can actually see the seams of the baseball as it's tumbling. The other thing that we did is we worked with another company to do optical tracking. And instead of your regular couple of camera machine vision stuff, these things are essentially two eyes that stitch together two panoramas and see stereoscopically so that we can see the nuances and reaction time of where the players are and where the position is. And it's amazing to just generate this data, to give you a little bit of the scale, like how big are we talking? We're talking about 17 petabytes of raw data produced, huge amount of data. And all that data has got to be turned into something that's actually useful to the customer. So it's not just the raw data points, it's what do we do with those data points? How do we turn it into a narrative of the game? And so we looked around, AWS was a clear choice. They were the only ones who had the technology, the only ones who had the scale, the only ones that we could take advantage of the ability to burst during the season when we need a lot of this, and then come down when we don't need to in the off season. And on top of it, we also realized that we were going to have an ever-increasing data custodianship of this data as we went season over season. And so it's not just a one-time shot. Every year, we're going to add more data to this repository, and how are we going to keep up with that growth? So Amazon, again, clear choice. So let's take a little, I'm just going to give you a real high-level overview of how the architecture is actually going to work. And so we'll take a look at this. And really where we start is by capturing that information that we showed using the systems that we did, transmitting it over our private network we've built out to all the ballparks, and then using DirectConnect, getting it into the Amazon ecosystem. So you see things in here. And again, there's not going to be a quiz, so I don't expect you to read this slide. But we do have things like Kinesis and Redshift, and a lot of the stuff that makes Amazon great. Ship the data up there, run a lot of these very fancy expert models, then create it, consumer products, and then get it out to the customer. So whether it's on their cell phone, or their PlayStation, or their Amazon Fire, we really want to get the most real time and most comprehensive data set that we possibly can. So all this sounds cool. It all sounds cool. It all looks pretty cool. How's it going to change the game of baseball? And so I bring your attention. We're going to roll this in here. Game 7, World Series, Kansas City versus the Giants. This is a play you've probably seen if you're a baseball fan. Maybe even if you're not a baseball fan, you may have seen this play. And we'll take a look at it right now. But we're going to break it down using StatCast. And so 90-mile-an-hour pitch, unbelievable hit here. The ball comes off the bat at over 100 miles an hour. And what everybody talks about is what happens next. Right here, with Joe Panic, sick catch. This guy's amazing, right? He hits 12 miles an hour in 18 feet to make that play. So he's reacting faster than the speed of sound for that ball getting to him. That's how fast he gets there. Makes a flip, turns two, out. Changes the game, right? But I want to actually change the narrative. Let's actually take a look at what Hosmer did. So first off, he slows down. He gets to 12.9 miles an hour. And he doesn't think that there's any shot Panic's making this play. So he's trying to read the situation. Now he knows that he needs to turn on the gas. So he hits just under 21 miles an hour. And if we take a look right here in the super slow mode, we can see that he's out by 200s of a second. Narrowly, right? So the question becomes, what happens if he actually runs it out? So those of you who play baseball, coach, baseball, see literally, don't slide into first base, run. If he runs and keeps that top speed, getting no faster than what he accelerated to, this play looks totally different. I don't actually have a simulation because it didn't happen, but he is safe by a foot. So the guy goes from being out for two 100s to being safe by a foot, proving the baseball added to that. You don't slide into first base. Right? So hopefully we've given you a little taste of what we're doing here, trying to show you that what we're trying to do with baseball is give the fans insight to the game, show how tremendous our athletes are. And we're doing that with Amazon. Stackcast is something that we would have had a really difficult time figuring out how to do any other way except leveraging the cloud technologies. And while there's a lot of partners that we work with, there's absolutely no question in our minds that Amazon is the number one cloud provider and technology that we've ever thought leadership that we could have ever chosen. Thank you for your time. Thanks, Joe. That is super cool. Really neat. And as a baseball fan, thank you. So that was a great example of how a company and an organization is transforming the customer experience. Every company in the world has to keep transforming their customer experience and their business to remain competitive. And while having a lower cost structure is an example for transformation, it's typically not the main driver transformation. The main drivers are typically agility and innovation. And the cloud enables that in a very significant way. It's not just that you have this very large, fully featured technology infrastructure platform at your fingertips. So you don't have to reinvent the wheel when you're building applications. But it's also how fast you can deploy servers if you want to try and experiment and do something new. So in the old world, when you asked engineering teams how long does it take for you to get a server to try and experiment or put something in intake, you get answers like 10 to 18 weeks. And if you think about that, it's maddening. And it's demoralizing. And it means that most people inside the company don't think about trying new experiments and innovation, because they figure why bother. It's never going to happen. Well, in the cloud, you can spin up thousands of servers in minutes. And so that completely changes. When you couple that along with this big technology platform at your fingertips, it means that you can go from idea to launch in record time. And what it also does is it totally changes inside your organization the amount of shower cycles you get from your employees thinking about innovation. Because they know if they have an idea, it's worth it, because they have a chance to try it and see if it works and then expand it. So it totally changes how companies think about innovation and it expands the group of people inside a company that are thinking about new ideas. And if you think about what's happened in the last few years, you look at some of these longstanding businesses that have been completely disrupted by new companies in a very quick amount of time who've built on top of the cloud to do so. It's pretty interesting, right? You look at what Airbnb has done in hotels. Or you look at what Spotify has done in the music industry. Or you look at what Dropbox has done for consumer storage. These are all businesses that are built on top of the cloud in AWS. They really disrupted longstanding existing businesses. And so it means, as an enterprise, you don't have the option of moving slow anymore. If you want to remain competitive longterm. Otherwise, you end up falling into that death spiral that ultimately leads to extinction. You have to be able to move fast and be agile and innovate in this environment to be competitive. And I'm going to bring up, as our next guest speaker, somebody who works at a company who is invented in a lot of different areas, a lot of different segments, and they have a very ambitious plan to revolutionize the way that health care is provided. Here's my pleasure to bring up the Chief Executive Officer of Health Care Informatics Solutions and Services at Phillips, Yeroen Toss. Great to be here. Let me start by introducing myself. I'm Yeroen, and I have hypertension. And so, my two brothers and my daughter has juvenile diabetes and my dad is recovering from cancer with major complications. And yet, I believe we're a normal family. And if you think about it, many of the people here in the room have similar stories. I would say 50% of the population is already suffering from a chronic condition. And actually, I don't want to bring you down here because I really think that as we are here and you look around, one in every two males will get cancer in their lifetime. One in every three women will get cancer in their lifetime. But the good news is that with the technology that we're going to discuss today, we're going to actually allow these people to live long and fulfilling lives. So technology is on the cusp of really making dramatic, really dramatic change in the way we provide healthcare. So we're no longer patients. We're consumers of health. And the best way to illustrate it is to talk about this real-life case of Eddie. So Eddie, we identified, was at risk. So a couple of factors indicated that he might be at risk for prostate cancer. So we asked him to come in. And we measured his blood. And we found out that there was an indication of a high risk of cancer. So we actually measured that blood at the point of care. We interpreted it at the point of care. So we could tell Eddie what was going on. We could give him all the information, then, about his condition. So we asked him to come in. We took an MRI. And we got confirmation that there was actually a lesion. So what we did next is we gave Eddie the connection of all the other people that had prostate cancer and rolled them in a community. We took the MRI. And we actually then took a needle with real-life guidance through ultrasound and took a sample of the tissue. We took the sample of the tissue. And we immediately on the spot diagnosed the type of cancer through digital pathology, which is actually a digital microscope. Then what we did next is we actually took a DNA sequence. And now we had a very, very precise diagnosis. Eddie's reaction was, I got to get rid of it. I got to get rid of it, which is a normal reaction. But then we showed to Eddie that he could actually live with it. But he has to monitor it. And he has to adjust his lifestyle. He has to continue to be monitored because the alternative was actually worse. If you do radiation, you have a lot of complication. So what am I talking about? I'm talking about capturing real-time data, interpreting it in real-time, engaging the patient, and fitting the diagnosis and the treatment precisely to that patient. And then using the data to actually be able to compare the data with others. So what we're doing now, we're allowing the data to be shared, in this case with cancer centers around the world. So if we see something out of the ordinary, then we can immediately share it. So it's pretty amazing that today you can actually take all that information. But it's not easy, because I tell you, digital scan is 500 gigs. A genome sequence is 500 gigs. And then we have all the data that comes streaming in, streaming in. So how do you deal with those data? And then if you look at what we're doing at Philips, already today we have 190 million people in the hospital, in the intensive care unit, where we stream their vital signs to monitor them in an intensive care unit. Now what if we can take these algorithms and apply it to 50% of the population that live with those chronic diseases? And you've seen all the wearables. We actually talk about implantables now that allows you to wear a patient monitor any time. We have 4 million people that have lung disease and that have ventilators. We're connecting those ventilators. We can help people control their breathing. We have a thousand seniors with multiple chronic conditions every day being monitored at home so that they can stay together if it's an elderly couple and we can help them. So tons and tons of data, or should I say exabytes. So today already we have 390 million studies, imaging studies that we manage on behalf of our customers. Now what a treasure trove of data. What if we start applying deep learning to that? What if we can start comparing all these specialist notes that we collected around this? So what we're doing is we're creating this Philips Health Week digital platform where we combine that data. We combine the data from the studies, from the images, from the electronic medical records, from the input that we get from our patients and we combine that so that we can allow people to make it actionable. Because data alone is not enough. It's got to become actionable. It's got to become actionable for patients. I need to take my medication on time. Hey, I have chest pain. I need to reach out. It's got to become actionable for the care team. The guy has chest pain. I look at his vitals. We need to triage health here. So decision support actionable. If we have these thousands of markers that we have to interpret in real time, we need the intelligence in real time to deal with that. And it's got to be a great experience. It's got to be an experience for the consumer. It's got to be an experience for the care team that takes care of the consumer and the specialist, the oncologist, the pathologist, the radiologist. So that's what we're building today. And we're doing that together with Amazon because we're talking about adding a petabyte a month. A petabyte a month. And we're doing that around the globe in 100 countries. So there's only one company that we can do this that gives us reliability and the scale and the performance. And we're opening this up. We're opening this up in the next quarter to allow all of you to work with us to create the reinvention of healthcare to help us deal with all these consumers like yourself, like your friends and family, and these millions of people out there that need our help to really, really change the way we provide care. Together with Amazon, we're going to do this on a global scale and we're going to save lives together. Thank you so much. Thank you. Thank you, Jeroen. That is very exciting. Really could change the world for a lot of people. We're very honored to be partnering with Philips on this adventure. So we talked earlier about the new normal comprising companies of all sizes deploying to the cloud by default and also companies having at their fingertips this very broad, fully featured technology infrastructure platform so they can move more quickly. One of the questions we get asked a lot from customers is is there hope for a new normal in the world of relational databases? And for customers, the old world of relational databases is not a pleasant place and it's pretty rare that we meet enterprises who don't ask us in some form, can you help us shift from what we're doing right now to something that's a little bit more customer friendly? And the reason is because these old world relational database offerings are very expensive. They're proprietary. There's a high amount of lock-in and they have punitive licensing terms, not just allowing very little flexibility and moving to the cloud in the way that customers want but also in the auditing and fighting of their customers. And so it's why you see so many enterprises and companies trying to figure out how they can move as many workloads as possible to the more customer friendly and open source database engines, things like MySQL and Postgres. The challenge is to get the type of performance in these open source database engines that you get on these proprietary databases is really hard. It's not impossible, but it's hard and it takes a lot of work. And so the obvious question that we got asked a lot was is there something you can do to help solve this problem for us? And so I am very, very excited to announce the launch of a new, a brand new database engine called Amazon Aurora. And Aurora means a new dawn. Aurora means a new dawn. We've been working on this for three years and this offering will be fully MySQL compatible but it'll have five times the performance of the typical MySQL implementation. It's at least as available and durable and fault tolerant as the enterprise editions of the proprietary commercial database engines and high-end sands. It's highly scalable, it's highly secure, it's available through our relational database service, RDS, so you get all the same management capabilities that you get with our other four engines in RDS and it's a tenth of the cost of the leading commercial database engine solutions. There's been a lot of invention in this database engine and I'm excited to bring up the general manager of this effort to give you a little bit more detail. I welcome honor of Gupta. Thank you, Andy. I'm really excited to talk to you guys about Aurora. As Andy said, it's a project we've been working on for a number of years now in secrecy. So databases have been around for a long time and when they first got started, they're really pretty innovative, introducing things like SQL and transactions that we all take for granted nowadays. But they were expensive, monolithic software running on expensive mainframe hardware and they were managed by an expensive, high priesthood of DBAs. But back then, that's the way all software was written. The problem is 40 years later, databases are completely ubiquitous, but they're still built around that same mainframe mindset. They're still super complicated and they're still super expensive. So a few years ago, Andy challenged us with this question. Guys, there's got to be a better way. In a world where AWS exists, why would you build a database that way when you have AWS scale, AWS services, and AWS cost structure? So we went back and thought about it. And you know, the guy was right. There is a better way. And it's really the same way that you guys have been building your own services and your own applications. You don't build monolithic software anymore. You do service-oriented architecture. You use multi-tenant scale-out components. You build it on AWS itself. And you use best-in-class AWS services like S3, EC2, VPC. And you make them self-managed. That's what we've done. So Aurora basically moves about half of the database out of the process. And there's a brand-new log-structured storage system that's scale-out, multi-tenant, and optimized for database workloads. And it's integrated with a bunch of AWS services like S3 in this picture. And with that, even though we have all of those changes, we're completely compatible with MySQL 5.6. And we let you migrate your data in with just a few clicks, or for that matter, migrate it back out. And with that set of changes, we got a bunch of big benefits. With Aurora, you can run 6 million inserts per minute, or 30 million selects. That's a lot faster than stock MySQL, running on the largest instances available from AWS. Whether you're doing network IOs, local IOs, or no IOs at all. But Aurora is also super-durable. We replicate your data six ways across three availability zones. And your data is automatically, incrementally, continuously backed up to S3, which, as you know, is designed for 11.9's durability. And it's highly available. We tolerate transparently disk failures, az failures. And it's self-healing will automatically repair those failures in the background. Crash recovery takes seconds. There's no redo log replay. The database cache survives a restart. There's no cache warming required. Your read replicas are all available as promotion candidates, and there's no data loss. These are very high-end features that are only available, in some cases not available at all, in the highest-end commercial engines. But all that said, Aurora is also low-cost. For an R3.large, you're paying just 29 cents per hour. With no upfront capital expense, no commitments, we try to earn your business every hour of every day. So we're super-excited about it. We think it's going to be a game-changer. The preview starts today. And I'd love for you guys to try it. Have a great conference. Thank you. Thank you, Anurag. In case it's not abundantly obvious, I'm really excited about putting this in your hands. If you think about the power and the capability that the AWS platform gives customers, one of the questions that you constantly have to ask, we constantly ask, our customers ask is, what are the tools and capabilities we're giving customers to harness the power and functionality of the platform? And the service we just talked about, Aurora is a good example of that, where we've built a relational database from the ground up with the cloud in mind. It's why we're able to provide some of the features and benefits that Anurag just talked about. But another area that we've been asked a lot about and we've thought about is, how do we help developers and teams speed up their software development lifecycle? And this is an oversimplified diagram of the software development cycle. And typically, engineers will develop code and commit it and then they'll do all kinds of building and testing and integrating and they'll ultimately deploy it. And then they'll monitor the performance of the application and then look at the logs, look at the results, figure out what's working and then build the next thing. And it turns out that that's a flywheel that you can inject energy into that allows companies to be able to build more quickly for their customers. And so customers have asked us and challenged us to think about, are there ways that we could help accelerate this cycle even further? And it turns out, we didn't have to look very far to come up with some ideas. So we looked internally inside Amazon at the tools that we've been building over the last 19 years or so to allow Amazon to move quickly. And the first place we looked was the area of deployment. And we have a deployment service that we call Apollo internally. And we have inside of Amazon, we have a very talented set of engineering teams that have very high standards. And they rarely agree on any one component that is truly superlative. This is one of the few they agree on. Apollo is a tool that everybody loves. And when people actually leave the company, one of the things they say they miss the most about the company is Apollo. And to give you an idea of how much use it's getting in the last 12 months, we've pushed 50 million deployments through Apollo. That's roughly 95 per minute. So this is a service that's gotten a lot of battle testing and a lot of usage. And so today, I'm excited to tell you that we're providing this experience to all of you, our customers, in the form of a new service that we call AWS Code Deploy, which is a fully managed, high-scale deployment service for you that's available today. The capabilities you have in Code Deploy are that you can deploy, you can choose to deploy to all of your instances at once if you want to, but most who have large amounts of instances, what they really want to do is deploy to groups of instances to make sure the deployment is working out. So we allow you in Code Deploy to do rolling, upgrades, rolling deployments. If it turns out that there's a problem with your deployment, we have automated health tracking that detects that and stops the deployment automatically. And it makes it really easy for you to choose to roll back if that's what you want to do. We have a centralized place where you can view all of the activities in the deployment and see each one that's successful or not successful or that's different from when you anticipated if you then again want to stop and make some kind of change. Code Deploy will work with virtually any programming language and it also will integrate with your entire tool chain set so you'll be able to use the same tools you've been using. It's available today, it's free to use, we're really excited to give it to you and put it in your hands. When we were talking to customers about Code Deploy and sharing what we were thinking and then giving them access to early versions of it, a lot of customers said, this is pretty interesting, pretty cool. How about some other things? There's some other ways you can speed up our software development cycle. And we have two other areas that we're making announcements about today, services that will be available in early 2015 that I'll share right now. And the first of these two that I'm going to talk about, again, we didn't have to look very far to come up with this for customers. And that's because over the last few years at Amazon we have really gotten religion about continuous build, test, and deployment. And that's because if you've done a lot of software development everybody knows when you get to the very end of any kind of complex application and you haven't really been building and testing and integrating the whole time it is very challenging to get a candidate that's ready to launch. And when there are problems it's actually pretty difficult to decipher where the problem was and what caused the problem. And so it really changed the way we started building applications to continuously build, continuously test, continuously integrate. You know every incremental time you've committed something and you try to integrate it what the problem is so you can fix it before you get to the very end. And so that's the first of these new services is AWS Code Pipeline. And what it is is a continuous build and test and integration service and it enables you to have repeatable automated integration. You can take code from any repository and spin up any kind of integration policies that you want or tasks that you want. It's got workflow modeling and visualization so you can better see what you're integrating and what's working and what's not working. It integrates with all your existing tools that you use. And as I mentioned earlier, this is used very pervasively inside of Amazon. Over 80% of development teams are now using pipelines and what's now Code Pipeline for you. So this is the first of the two. The other is a service that we call AWS Code Commit. And so this is a managed code repository in the cloud. It's highly durable. It's highly available. It allows you to closely locate your code to where it's actually going to be deployed to for low latency. And then because we built this in the cloud, we were able not to impose any of the size limits on files or repositories that often customers are subjected to. So this set of services will integrate with existing partners of AWS and providers that you use. And it arranges from GitHub and Chef and Puppet and Elassian and Jenkins to environments like Visual Studio and Eclipse that you'll be able to use these services right out of the box. And this set of services is designed to work together. The reality is that when teams find that they're able to deploy code in a repository that's low latency and very reliable and that's managed for them and then build and test on a consistent basis and then deploy in a highly flexible way, they have much better success in deployments. And when companies have better success in deployments, they deploy more frequently, which means more features and capabilities for your customers. This is, I think, a really significant deal for our customers and we're really excited to provide it to you. So if I flash back a few years ago and we talked to enterprises about the cloud and AWS, security and compliance were blockers for them for moving the number of workloads that they really wanted to move. And another component of what we see as the new normal now is that security and compliance are becoming reasons that customers are moving to the cloud. And you see that in lots of different ways. If you look at the amount of certifications that AWS has achieved and secured for its customers over the last several years directly influenced by what you told us matters most, it's been a real enabler for enterprises to move. So it's SOC1 and SOC2 and SOC3 and ISO 27001. You can be PCI compliant in AWS. You can be HIPAA compliant on AWS. We have a number of public sector certifications like FISMA and ITAR and FedRAMP and the Department of Defense. And then just last week, we launched our latest certification, which is ISO 9001. And ISO 9001 is primarily for healthcare, life sciences, medical devices, automotive and aerospace. They all have very demanding quality mandates for them to be able to deploy their applications. And ISO 9001 enables them to meet those quality mandates in AWS. And this was inspired in part by working with our healthcare customers who really wanted to be able to meet GXP requirements. We're the only cloud provider that has this and we're really excited about what it enables. Now, if you think about security and you think about customers that care deeply about the privacy and security of their data, those companies encrypt their data. And it's why we've built a number of encryption and key management solutions in AWS over the last several years because we know that's important to our customers. And we have some customers who are comfortable with us managing the keys and so they use the encryption in services like S3 and EBS and RDS and Redshift. And then we have other customers who want to manage the keys themselves and they can either bring their own keys to services like S3 or they can choose to store those keys in a service that we have called CloudHSM, which are hardware security modules. And if somebody unauthorized tries to access those keys, it shuts down and the keys are destroyed. But when you talk to customers about these solutions, what they tell you is there are a number of challenges regardless of which way you go. For those that want us to manage the keys, they'd like more control and they'd like more visibility over those keys. And for those that choose to manage the keys themselves, what they say is, look, it's actually quite difficult to store those keys consistently in a safe way. And key rotation is really hard. I don't know how many of you tried to rotate keys on a complicated application while it's live. It's challenging to do it without bringing the application down. And so what happens is most companies don't do that key rotation. Or they say, look, it's actually hard to have visibility into who's accessing these keys across the company. And for those that store those keys in HSMs, the HSMs are expensive. And it's typical that you don't have very high utilization of those boxes. So it's kind of an expensive solution. And so we tried to find a solution that was the best of those two worlds. And today, we're announcing that it's immediately available for you a new service called the AWS Key Management Service, which is encryption and key management and compliance made easy. And what this allows you to do is you can encrypt in one click from the AWS Management Console, or you can choose to do it programmatically through the SDK we provide. You have a central place to create keys, disable keys, view all those keys, or set policies on keys. You have enforced and automated key rotation, so you actually are rotating your keys. And then you have visibility into any accesses to those keys through AWS CloudTrail. The keys are stored and used in HSMs. They're integrated with all of the key data store services in AWS. There's S3 and EBS and RDS and Redshift. And then it's built as a global, highly available, highly durable solution. And this really matters if you think about it. If you have an application where you've got encrypted data and the application has to decrypt that data to be able to display whatever the data is that you're trying to present, if you don't have a key management service that's highly available, it will take your entire application down. So we're very excited about this service. We think it will give you a capability that solves a problem that we've heard consistently from customers with the two different ways they've been trying to do key management and it's available for you today. So encryption is one very important component of security to customers. But what we've heard repeatedly from customers is we need to be able, as we're moving all of these workloads to AWS, we need to be able to meet the same compliance standards that we've set out for ourselves. And so one of the reasons that we announced last year at ReInvent, the new service that was AWS CloudTrail, and CloudTrail for those who don't remember, is the service that logs and tracks all of the API calls made to your AWS resources. So who made the call? Where it came from? What resources they tried to impact? And what resources they successfully impacted? And you have companies like FINRA and Siemens that are using CloudTrail for sensitive workloads they use in AWS. But frankly, the vast majority of our enterprises are already adopting CloudTrail. It was amazing to us how fast the adoption happened. But our customers want to have visibility into more than just the APIs. They want to know what resources they have in AWS and how they're related, and they want to be able to see what happens if they make changes for these resources that are part of various groups they've set up. And the traditional CMDB and ITIL solutions were built before the cloud came about. And so they really have a high emphasis on locking things down and making change hard. And enterprises don't want to make changes willy-nilly, and they don't want to make changes that they don't need to make. But the biggest value, the primary benefit of the cloud is agility and speed. And so you don't want a tool set that prohibits you from being able to take advantage of the power of the cloud. And so we're introducing today a service in preview called AWS Config, which is a new resource dependency and auditing service for you. What this service allows you to do is it lets you have full visibility into all the resources you have in AWS. So if somebody says, tell me everything you're running in AWS, you can answer that question with a service. If somebody says, do we have any instances or do we have any resources that aren't running in a VPC, an Amazon Virtual Private Cloud, you can answer that question. So it gives you full visibility into all your AWS resources. And then it allows you to understand the relationships between all of those resources and the dependencies. So one of the challenges that we all have is when we make changes, say we're going to change security groups, is what's the blast radius of that change? What might be impacted? And oftentimes we find that out after we've made the change. AWS Config allows you to know that before you make that change. So visibility into all your resources, relationship between those resources, and then it allows you to see any change to configuration that's been made. And so you can subscribe to a stream that you either consume programmatically or that you consume through the UI in our console where you can see by the hour, by the day, any configuration change that was made. And you can import that data back into your own environment so you can compare that to what you're running on-premises as well. And all of it is tracked in CloudTrail too. Like all the other services we're talking about, these are going to be supported by our partners and integrated into their solutions. So again, you can use the same tool chain that you used to using. So encryption really matters to customers and having flexibility in how they do what matters. Having the right visibility into their resources and changes to their resources really matters. The other thing we hear a lot from our customers is I want to move large amounts to AWS. I'm actively moving large amounts to AWS. However, I also want to make sure that I can have a standardized way that my employees are deploying to AWS. I want everybody in the company making up the way that they deploy to AWS. And this is why service catalogs are very useful in enterprises. And in fact, Forrester said, the cornerstone of service and delivery automation is service catalogs. And it allows enterprise administrators to set up how they want resources deployed, who can deploy them. They have access control along the live dimensions. And then they hope to have visibility. Some have the visibility, some it's a little bit harder. And not surprisingly, as so many enterprises have been moving to AWS, we have heard repeatedly from our customers, we would love a service catalog. Please bring us a service catalog. And so it's my pleasure to announce today coming in early 2015, the AWS service catalog. And this will allow enterprise administrators to create portfolios of products set up in the configuration in the way that they want them deployed. And then they can make them easily discoverable for employees at their company on a portal that they host. And then it allows them to have fine-grained access control. They can have access control on the individual, on the group, on the department, or on the cost center so they can meet the compliance needs they have in different businesses. And then again, they have that strong visibility as all of the activity is tracked in AWS CloudTrail. So this set of services is very important to our enterprise customers. And what we've found is that over the last year or two, what our customers have said is, look, I get that the individual building block infrastructure services are secure. I've looked at the architecture. I've talked to your security guys. I've read the white papers. I've seen the track record. I get that those building block services are secure. But you have to help me find a way to make my organization secure around those services and how they use those services. And so giving customers a better way to do key management encryption, full visibility into their resources and the dependencies of those resources and any changes to that resources and a way to have a standardized way that they deploy to AWS is really important to our customers and that's why we've provided these three services. So one of the questions we get asked a lot is how are enterprises moving to the cloud? Are there any patterns? What's the sequencing that you typically see? And not surprisingly, there is no one-size-fits-all way that companies are moving to the cloud. But we have seen a number of patterns that I thought I would share with you. Typically, as enterprises are moving to the cloud, the first set of workloads, there's a couple of waves that they move in. The first wave typically includes development and test. That is a no-brainer. And we have lots and lots of enterprises running their development and test environments on top of AWS. For example, as you can see, the Tokyo Stock Exchange is running Oracle and Dev and Test on top of AWS. Lionsgate is running SAP and SharePoint in Dev and Test on the way to production on AWS. You can see Fidelity now is running a lot of Dev and Test environments on top of AWS. So Dev and Test, and then the second workload that moves in that first wave are altogether new applications. And that's also a no-brainer because there's no legacy. There's no migration challenge. And you get to take advantage of all the benefits of the cloud right from the get-go. You heard Major League Baseball talk about StatCast. GE has built an innovative new crowdsourcing innovation application that connects people and products and materials and simulations. Siemens has built a new healthcare application that makes it easier for doctors to provide personalized care to their customers. Boeing has built their supply chain application for all their providers on top of AWS. So that first wave that typically moves to the cloud is Dev and Test and altogether new workloads. The second wave encompasses three other categories of workloads. The first is websites and digital transformation. This is a pretty obvious one that goes pretty quickly. They're relatively straightforward to move. Again, lots of examples you can see behind me, but Unilever is running about 1,700 websites at this point on top of AWS. They've migrated several hundred that they did in just a few months. The Navy is running a number of workloads on top of AWS, a number of websites, and saving 50% in the process. The Singapore Post has built a brand new e-commerce business that they're running entirely on top of AWS. That's one of the fastest businesses in that organization. So you see a lot of websites and digital transformation in that second wave. You also see a lot of analytics. It has never been easier and less expensive to collect, store, analyze, and share data than it is with the cloud. And that's why you see so many more companies holding onto their data, but actually doing something with the data, analyzing it and taking action on it. You see NTTDocomo is running a four-petabyte data warehouse on top of Amazon Redshift. The Financial Times is using AWS to do analytics for their 450,000 subscribers. The queries they run on AWS are 98% faster at one quarter of the cost. You can look at companies like Telinks. Telinks is a company that does... It's one of the largest insurers in Germany. And every company in Germany now has a requirement to run what's called a Solvency 2 calculation. And what it is, is that you have to take all of the customer policies and create mathematical calculations out of them and then assess if all those debts got collected once, would they be solvent? And so they run those calculations now on top of AWS. They're 75% faster than what they were able to do on-premises, 60 to 70% less expensive, and they're saving 8 to 9 million euros a year. So you see websites of digital transformation, analytics, and then the third wave of applications you see in this second wave are mobile. And most companies have employees now that have three screens. They have a phone, a tablet, and a laptop. And then most of their end users want to consume those applications via mobile devices. So again, you see our customers that are building new mobile capabilities on top of the platform. And they range from top to top motors, which is able to transmit all of their GPS coordinates so they're truck fleet on top of an application that they built on AWS in real time. Or you can see the Xiaomi, a very large device provider in China, runs their mobile app store on top of AWS. Or you can see Vodafone in Italy has built an application that allows customers to top up their payment accounts from multiple vehicles. Or you can see Quantus Airlines, which built a brand new application for their flight attendants to be able to provide more personalized service on the plane about their customers. In that second wave, you see lots of websites, lots of analytics, and lots of mobile. And then the third wave is typically the more business critical applications. And if you think about it, this makes sense. These have the most amount of dependencies, the most amount of migration pain, and you have to do this very thoughtfully. And again, lots of examples here. I'll just name a couple. Connie Nast has moved their legal, their HR, their finance applications all to AWS. Or you can look at what happened in the Netherlands. The Dutch national bank about a year ago said that it was safe for financial services companies to run on top of the AWS platform and enabled Rovaco of Bank of the Netherlands to build their entire retail banking platform on top of AWS. So you see these waves of migration and typically the sequencing that we see. But what you're seeing increasingly now are two other trends. You're seeing companies migrating entire data centers now to AWS. And what they're doing is they're looking in the future and they're saying, well, I have a data center that's coming up for lease in 14 or 18 months. Or that I'm going to have to do a significant technology refresh in. Maybe I don't have to do that. Maybe I can figure out a plan and move all those workloads to AWS. And again, you're seeing more and more examples of this. Connie Nast moved their primary data center to AWS in just a few months. They did it moving all their most sensitive workloads and applications there. They're saving 40% in terms of cost and they're getting better performance. Or you can look at Hess. Hess decided to divest from some of their downstream businesses and when they were providing all the applications to the acquirer, they put them all in AWS, made sure they could run in AWS and literally handed over the keys. Or you see News Corp. News Corp is in the process of moving 75% of their workloads to the cloud and to AWS. And in the process, they're going to take the number of data centers they use from 45 to 6. So lots of companies migrating entire data centers now. And then increasingly, lots of companies moving all in on top of AWS. You know, when Netflix announced a few years ago that they were moving all into AWS, it was a pretty bold and courageous decision at that time. I would say today, it's not as bold a decision. It's actually becoming more than new normal. And in addition to Netflix, you see companies like Sun Corp, which is the very large financial services and insurance company moving all into AWS and the Kempinski Hotel Group in Germany and the real estate enterprise company GPT Time, Inc. is moving all into AWS. You see the healthcare services provider, MDR moving all into AWS. You see NEPON Express in Japan in the process of the next few years moving everything to AWS. And the same goes for ISVs. We announced, along with N4 back in March, that they were moving their underlying platform to AWS for their cloud delivery. And you see the same thing has happened with companies like Splunk and Acquia and Informatica and the large security company Asset, IMS Health, PEGA. This is a trend that is accelerating at a very fast rate. Now, customers still ask us, can you give us examples of sensitive workloads, large amounts of sensitive workloads that are moving to AWS? And we got to ask that question a lot more a few years ago, but we still get asked it and it's a good question. And the tenor of that conversation really started to change a year ago, not just with the amount of companies that have been announcing that they're moving all into AWS in lots of industries with very sensitive data, but when there was a significant amount of press that the CIA chose AWS to run their infrastructure on top of and a lot of our customers would say to us when we meet with them, well, if it's secure enough for the CIA to use, it's probably secure enough for us too. And that has completely changed a lot of these conversations. I'm really excited to reveal the next company in a very important industry, financial services, with lots of sensitive data that's moving all their applications to AWS. This is a company that's been really a pleasure to work with over the last couple years. I'd like to welcome to the stage Taylor Stansberry, who's the senior vice president of CTO of Intuit. I'm really delighted to say a few words about how Intuit is using AWS. You may know Intuit for products like TurboTax and QuickBooks, but our mission is to improve our customers' financial lives so profoundly that they cannot imagine going back to the old way. We serve consumers with products like Quicken and Mint, small businesses with products like QuickBooks and Payroll, and accounting professionals with products like Lassert, Protax. Intuit was founded in 1983. We have some 8,000 employees now, of which 3,000 are engineers. We serve 50 million customers. We gain 4.5 billion dollars of revenue a year, of which about 3 billion come from mobile and online services and applications. Mobile and online is growing rapidly for us, and we expect it to continue to do so over the coming years. And we are also currently operating actively in six countries today, and we expect that to grow to some 20 countries over the next two or three years. Intuit is an awesome place for engineers with a strong culture of innovation, working hard on making complicated things easy for our customers. Now, here's where we started to engage with AWS. We had an application called TurboTax Answer Exchange, which helps our customers get answers to their questions about how they can file their tax returns. It's complicated because it tries to parse every question and feed it to people who have answered like questions before and gotten high ratings on those answers, and it runs at scale. But most of that runs since idle for most of the year until we get to the peak of tax season. It happened to be running primarily in a data center whose lease was coming up, and we did not want to renew it, so we needed to move it out fairly expeditiously before tax season, and we looked around at our options as to what was the best way to do that, and AWS came clearly to the fore. So we were able to move it in time, and the results were excellent. We were able to cut costs by a factor of about 6x versus having servers that sat idle for most of the year. We were able to get it set up in about a fifth of the time it would have otherwise taken to do a build-out and deployment, and our developers were able to move much faster. So building on the great results we had with TurboTax Answer Exchange, we moved a bunch more of our applications and services to AWS, and we realized a number of further benefits. For example, being able to deploy in data centers that Amazon already had it set up in regions that we wanted to serve. Last year we did 10 acquisitions, and about half of them were already running on AWS, which made the integration of those acquisitions much easier. And of course, developers love to work on AWS, so it's been a great recruiting tool for us. Today we have some 33 applications running in AWS, 26 services, and eight enabling tools. Just a few examples of those include the QuickBooks marketing site, which we were able to get set up and reach global scale in about a month. Also the TurboTax marketing site. We run our customer A-B testing and telemetry out of AWS. We have an application called TextWeb, which helps millions of customers in India on feature phones over SMS get web-like functionality, and AWS enables us to scale up and down rapidly when there are events like World Cup Cricket that drive a lot of activity. All of our mobile services are running in AWS. And most recently, I'm very proud to announce that we have moved Mint.com to AWS, all of the computing from Mint.com. Mint.com is a personal finance solution that serves millions of our customers and is free, and I highly recommend it if you don't use it today. Along the journey in our moving our applications and services to AWS, we've taken full advantage of all of the rich set of services that AWS has generated. Now, we take our customers' data very, very seriously. It's after all their data, and we're the stewards of that data, and it's our job to keep that data secure and to make sure that that data is used only for purposes that our customers would want. So we have worked very closely with the AWS team to help back additional services to further secure our data and, as you heard just now, we worked with them on CloudTrail, on the key management service so we can manage our own keys on VP peering and on others. In fact, we've been so pleased with our partnership with Amazon and their positive impact on our company that we named Amazon Partner of the Year. So what's next is that over the coming years, we plan to move the rest of our applications over to AWS. And we're doing this so that we can speed development so that we can innovate faster so that we can solve our customers' problems better. Thank you all very much. Thank you, Taylor. It has been really a pleasure and an honor to partner with Intuit over the last couple of years. And as Taylor mentioned, they have contributed very significantly to the services that we launched today. And they've also contributed to a lot of capabilities that we haven't released yet that we will release soon. That all of you will be able to enjoy. So, we just went through a number of different groups of workloads, including entire data centers and all-in migrations to the cloud and AWS. You know, sometimes people get confused and they say, well, wait, is what you're saying that you either have to move everything to AWS or nothing to AWS? And of course, that's not a binary decision you have to make. We understand that there are a lot of enterprises that have on-premises data centers that they're not ready to retire yet. And what they really want is they want to be able to run those on-premises data centers alongside AWS in as seamless a fashion as possible. And it's why we've spent so much energy over the last few years and will continue to do so to build capabilities that allow our customers to run in hybrid mode on top of AWS. It's why we built Amazon VPC, which really is effectively coordinating off a part of our network, allowing you to deploy resources into a VPC. You can set up subnets that are either only privately addressable or can be publicly addressable, and then connect via VPN to your own data center. So, VPC really becomes an extension of your data center topology or why we built Direct Connect, which is a private connection between your own data center and our region. It's why we've built Identity Federation and Access Control. It's why we've built AWS Directory Service, which we launched just a few weeks ago that allows you to connect your Active Directory or SAML Directory to AWS. We had a lot of customers who said, hey, look, I've invested a lot of time and energy into tools either with vCenter or system center. And what I'd really like to do is use those same tools. We'll be able to use AWS with them. And so we've spent a considerable amount of time building plugins that allow you to import and export VMs back and forth and manage those footprints with the same tools you're used to using. It's why we've built things like CloudTrail and Config, which allow you to import that data and use it in conjunction with your other data on-premises. So we've built a large number of these services that allow customers to run in hybrid mode. And we have lots of enterprises who are indeed running in hybrid mode. And I'm going to bring up to the stage now one of those customers from a company that's not only been successful for a long period of time, but has also reinvented lots of different industries. And it's our honor to work with them. I'm going to bring up now the Chief Technology Officer of Johnson & Johnson, Dan Zella. Good morning, everyone. It's my pleasure to be here with you today. I want to give a brief update on the amazing journey we at Johnson & Johnson are on from a technology transformation perspective. I'm going to start off by telling you a little bit about Johnson & Johnson, our operating model, and some of the market forces that we live with every day. So the Johnson & Johnson brand is world-renowned. People associated with things like Band-Aids, Tylenol, Johnson's baby shampoo, there's so much more than that. We're the sixth largest consumer health company, and pharmaceutical and bioinformatics company as well. Manufacturing drugs in critical cases, like cancer, diabetes, et cetera. We're also the world's largest medical device manufacturer and device company. In terms of our operating model, we're over 270 operating companies that operate in 60 countries across the globe. So truly a global company. From an operating model perspective, it's a highly decentralized model. Many of these companies would be rated as small to mid-sized companies that operate under the J&J umbrella. Many of them were through acquisitions. I would say upwards close to 50% came through acquisition, M&A activity. And again, it is through a highly decentralized operating model that we operate in, where there is a lot of autonomy that is giving to the local companies to innovate within the healthcare space. And lastly, we look at the market forces that we live in today. Healthcare reform has changed the game for everyone. With a drive for efficiency and transparency, it's not only impacting our customers, but it's impacting us as well, where there's more focus on more discussions on value, more discussions on transparency, more discussions on what should be on formulary, what shouldn't be on formulary. It's really changing the face and the pace and the need for crisp analytics and data. And you have market disruptors. Today, more than ever, the barriers to entry are lower than they've ever been. All it takes is cloud, mobile, digital devices, and this is particularly true in emerging markets, where J&J has a huge opportunity from a revenue growth perspective. But what worked years ago in those markets from a cost perspective and a technology footprint would work today. So we need to take a very different view of how to operate and deliver technology in those places. So before I talk about where our journey and where we are, I think it's instructed to talk about where we've come from. So, like any large global enterprise that has grown through acquisition that has a decentralized model, we have a heterogeneous distributed infrastructure of technology across the globe. It's one that was highly physical in nature, where low levels of virtualization existed. Provisioning was measured in weeks versus minutes, let alone that. Low utilization. So we had overprovisioning going on because people wanted to get their swing at the bat to make sure they were provisioned on time. Low financial transparency due to the fact that much of this is a capital intensive environment. And a high degree of friction in terms of exit and entry. And if you think about many of the small companies, they need that ability to experiment and get in and get out quickly, which with the capital intensive environment and highly physical was not really easy to do. So our business partners were asking us for something different. Quite frankly, better, faster, cheaper. So about a year ago, we embarked on laying down a new strategy. You know, on our first swing at that, we weren't bold enough. It was incremental. Yes, we were going to virtualize more to compute level. We were going to lower our technical debt. We were going to do some process improvements. And we just realized it wasn't bold enough. So we took a step back and indeed bold we are. We're going all in on a hybrid cloud strategy for J&J. Part of that being an on-premise cloud environment, which are in the midst of rolling out across the globe as we speak. We partnered with Amazon as well for virtual private cloud and public cloud capabilities. And we're on this journey now where it's all about collapsing the infrastructure, automating as much as we can, creating seamless interoperability with the data centers. I like to refer to the hybrid cloud as the borderless data center. And I think increasingly that's a model. All of us large enterprises are going to face as a reality where it's not going to be clear where the data center begins and ends. But what's going to be important is to maintain that operational control of visibility. It's something I focus on quite a bit. And with that, we're looking for fine-grained controls in terms of elasticity to enter and exit out of these environments. And where we can, we'll leverage SAS and pass. So we've partnered with Amazon, again, on our VPC environment. And today we have approximately 120 application workloads running in an environment. And over the next year, we expect to triple that easily. And as things like GXP compliance come along, which we, working with Amazon and others, are a voice and strong advocate for that to become a reality, that'll also open up new opportunities for us in life sciences. So one area is application. The other areas where we're focusing on, one is big data, which we've been at that for a while now and pretty far along. And the other is on the desktop, and I'll talk briefly about both of those. In the big data space, we have work with Amazon and the various tools that they offer to provide this big data simplified architecture patterns. And what the focus is on here is minimizing the time to insight. Less about gathering the data, less about provisioning. How do we get to the insights quickly? And we've had remarkable success thus far in taking structured, unstructured data, taking streams of data from our supply chain to create value in insights in a more rapid fashion. And we've done it in areas like physician compliance. We've done it in our supply chain area. We're taking real-time feeds from the manufacturing line to measure quality and efficiency. In the bioinformatics space, we're looking how to leverage the elastic compute capability of Amazon to do highly intensive compute operations, merging of data in complex models. Additionally, we're looking how we can analyze our current biologics or medicines that we have to create new medicines and have any stability to churn to the data and spin it up and spin it down has really been an amazing benefit to us and it's just growing, growing every day. Now lastly, we'll talk about desktop and work spaces. This is really our next venture here with AWS and one we're really, really excited about. It's one that we've had a large influence on, I believe, and AWS has been amazingly responsive to putting our security needs in place and our operational needs in place into the product. And so our strategy here is to bring the power of the cloud to the desktop, to decouple the physical device from the operating system, such that we can create more efficiency in how we provision, how we do patches, how we do M&A activity, how we promote self-service and BYOD. Our initial use cases are going to be for contractors and employees. Today we issue thousands of laptops to contractors who work for J&J at a significant capital cost. This changes that dynamics. It also helps us with level two support. So we're really, really excited about this opportunity. We'll also use it into zero desktop footprint space for whether assured workspace environments or people who are office workers. Our plan is to roll it out to over 25,000 desktops in 2015. So again, we're really excited about it. The economics are there. We're getting tremendous feedback from our users on not only usability, but performance around the globe. So the user interest and demand is there. The time is right. And we're really excited about doing that. And this was just a brief update in the various areas that we were working with AWS. And I have to say it's been a great experience for us as an enterprise customer work with AWS. They challenge us, we challenge them, and they deliver. Thank you very much. Thanks, Dan. J&J has also been an incredible partner in not only some of the services that we've delivered already, but in, again, several of the others that we're in the process of working on that we'll deliver to you in the coming months. So I thought I would close this session by talking about something we get asked a lot as companies are now on this journey of moving to AWS in the cloud. And the question goes something like this. As we think about investing in your platform over the long term, how does Amazon work? What's the culture like? What's unique about your culture that's important to us when we think about partnering with you? And there are a lot of things that I think are unique about our culture. But I usually point to three things. The first is everything we do at Amazon starts with the customer and then all the strategies and tactics move backwards from there. And I think there are a lot of companies who claim that's what they do, but very few act on it. I think there are companies, in fact, most of the large companies in the technology space tend to be competitor-focused and they tend to decide what to do based on what other companies do. And that can be a good strategy. It can work sometimes. You're typically chasing all the time. Or there are some companies who say, look, we're a product company. It's nice to hear what you guys think you need, but leave it to the experts. We know what you want. And that's also an approach that could work. It just happens not to be ours. We are a customer-obsessed and customer-focused company. Everything we do depends on what you guys tell us matters. 90% of our roadmap is driven by feedback we get from you in forums like this. It's why we love this week so much. And the other 10% is driven by things we hear you say, problems you want to solve, but we then have to think about what you really want and invent on your behalf to solve those problems for you. So number one is that we are a very customer-focused company. Number two is we like to invent and we like to pioneer. We hire builders and we hire inventors who are inspired by trying to figure out how to reinvent customer experiences in various businesses. And again, you can have lots of strategies you pursue. A lot of the large technology companies acquire most of their innovation. They've either lost the gene internally to invent or don't have the will to do it, and they tend to be fast-following. And that could be a strategy that works. But what I would say is in fast-moving industries and fast-moving shifts, and the cloud is the biggest technology shift in our lifetimes, it's usually a big advantage to work with the pioneer and the leader who has more functionality, who iterates faster, who has more customers, who has a larger ecosystem, who has more learning given what they've been doing and for so long with such a large advantage, and who had the vision for the space from the start to the services fit together as opposed to having to work with companies that acquire all these different pieces and try and make them fit together. So number one is where customer focus, number two we like to invent, and number three, we are extremely long-term oriented. And we like to think of the businesses that we build at Amazon as planting seeds that will hopefully grow into large trees down the road. You will never hear us or see us call you the last day of a quarter or the last day of a year with some kind of special deal to make numbers in that quarter. We are trying to build a business and trying to build customer relationships with you that outlast all of us. And I think an example that ties this together pretty well is something called Trusted Advisor. And this is a capability and a support function which is hundreds of programmatic checks on how our customers are using our platform. And when we see there are areas that you may not be optimized we proactively reach out to you to tell you. So an example might be we see security ports open and we say, hey, do you really mean to have the security ports open? Well, another thing we do is we look at your utilization of the resources in AWS and when you're paying for resources that are idle or that have very low utilization we proactively reach out to you and say maybe you should stop paying for those. And we've done that. If you look at the last couple years we've sent over 2.6 million of those notifications that have actually saved customers over $350 million. Now I ask you, how many large technology customers or companies call their customers and say stop spending so much money with us? Not that many. And when we first did people thought it was a little bit wacky. But if you think about what I just talked about it makes a lot of sense. We don't want to make money from you if you're not getting value out of it. We seek to reinvent the customer experience in every space that we operate in as a company and we're trying to build a long-term business that outlasts all of us in this room. So it makes a lot of sense if you understand our culture. I want to thank you for being here. I want to thank you for spending your time the next few days. And my hope is that you take this learning that you get the next three days and you take it back to your companies and change your business and change the customer experience for your customers. Thank you very much. Ladies and gentlemen, this concludes today's program. Please rejoin us tomorrow for Keynotes at 9 a.m.