Google Unveils Cloud Spanner and Price Cuts to Win Enterprise Workloads

谷歌推出Cloud Spanner并降价以争夺企业工作负载 谷歌认为，僵化的供应商锁定导致企业浪费了近一半的云预算，因而需要灵活的承诺机制和托管型全球数据库来加速关停数据中心。 本次主题演讲概述了Google Cloud在全球网络基础设施、数据库工程和开发者工具方面的广泛企业级投资。主要产品亮点包括推出具备水平扩展能力的Cloud Spanner、灵活的定价模式、Identity-Aware Proxy，以及Cloud Functions和托管数据服务等无服务器选项。演讲还详细介绍了G Suite向企业优先的转型，推出了Team Drives、Hangouts Chat和Meet、Jamboard，并与Apigee进行深度集成以赋能现代API生态系统。 在近期近300亿美元资本支出的支持下，Google Cloud通过定制设计的基础设施承载了全球预计25%至40%的互联网用户流量。 (0:03) Cloud Spanner作为一款托管型关系型SQL数据库推出，通过TrueTime原子钟将水平扩展能力与全球外部一致性相结合。 (4:41) 谷歌推出了企业安全解决方案，包括公开测试版的Identity-Aware Proxy (IAP)、正式商用（GA）的Cloud KMS以及Cloud Data Loss Prevention (DLP) API。 (33:10) G Suite通过正式推出Team Drives、Google Vault for Drive、Drive File Stream以及全面重构的Hangouts Chat和Meet体验，扩展了企业协作能力。 (87:21) 收购Apigee后，Google Cloud将API管理和互联业务平台定位为混合云采用与数字化现代化的核心组成部分。 (110:57)

At Google Cloud Next '17, Google leadership argued that modern cloud adoption requires moving beyond rigid infrastructure toward elastic, software-defined systems and integrated business ecosystems. Across compute, security, managed data platforms, and collaboration tools, the company positioned its own hyperscale stack as the blueprint for enterprise modernization. Global Infrastructure and Flexible Cloud Economics (0:03) Google detailed substantial investments in custom hardware and a private global network, including proprietary undersea fiber cables like the Unity cable and high-capacity links connecting the Americas and Asia. Urs Hölzle announced new cloud regions in the Netherlands, Canada, and California to bring Google Cloud Platform's footprint to 17 regions and 50 availability zones. Alongside this expansion, Google unveiled VM configurations supporting up to 64 vCPU cores and 416 GB of RAM, with larger footprints planned for later in the year, as well as the ability to attach GPUs to arbitrary VM configurations. Hölzle directly challenged industry cloud economics, citing research indicating that cloud customers waste an average of 45% of cloud spend due to inflexible commitments, rigid power-of-two VM shapes, and coarse billing increments. In response, GCP introduced committed use discounts offering up to 57% savings for aggregate core and memory commitments without upfront cash, alongside granular per-minute billing. Customer speaker Ashok Belani highlighted how Schlumberger uses this tunable compute and GPU infrastructure to process massive offshore acoustic datasets spanning months of raw acquisition into complex subsurface imagery. Globally Consistent Databases and Data Pipelines (7:51) A primary highlight of Google's data infrastructure was Cloud Spanner, a fully managed, globally replicated relational database combining standard ANSI SQL semantics and strict ACID transactions with horizontal scalability. Live demonstrations illustrated Spanner scaling instantly from 70 to 99 nodes and performing schema changes without downtime or query disruption on an 80-terabyte database processing up to one gigabyte per second of throughput. Google noted that the system automatically shards and balances data based on real-time access patterns. For enterprise analytics and integration, Brian Stevens and Greg DeMichillie outlined platforms designed to handle streaming and messy real-world data. Demonstrations combined Cloud Pub/Sub, Cloud Dataflow, and BigQuery to process over 15,000 real-time taxi telemetry events per second. The company also introduced Cloud Dataprep to address industry estimates that data scientists spend 75% of their time cleaning disparate data sources, using machine learning to detect anomalies and transform structures before running inferences via Cloud ML Engine. Application-Level Security and Identity Control (33:10) Google framed perimeter security as obsolete, identifying credential phishing as the primary security vulnerability facing modern enterprises. Urs Hölzle argued that software VPNs create an unmanageable administrative burden while leaving vulnerable internal networks exposed once breached. In their place, Google introduced the Identity-Aware Proxy (IAP), an access-control layer built directly atop GCP's global load balancers to enforce device-level and identity policies without client-side VPN software. A live demonstration showed an on-premises Oracle E-Business Suite instance secured instantly using IAP and physical hardware security keys. At the data layer, Google introduced the Data Loss Prevention (DLP) API to prevent personally identifiable information (PII) leakage. A live support chat demo illustrated the DLP API scanning chat logs and uploaded image attachments, automatically detecting and redacting names, phone numbers, and social security cards across text and visual media. Serverless Developer Velocity and Modernization Pathways (43:47) Brian Stevens described enterprise migration through three distinct pathways: lift-and-shift, cloud-native development, and targeted analytics adoption. To accelerate early transitions, Google announced a hypervisor-agnostic live migration tool operating directly from the Cloud Console across bare metal, Hyper-V, VMware ESX, and other public clouds. Lush Chief Digital Officer Jack Constantine highlighted migrating 17 global e-commerce sites to GCP in 22 days during their peak December trading window. To decouple modern services from legacy backends, Google emphasized Apigee and serverless tooling. The newly launched public beta of Cloud Functions allows event-driven automation across GCP services, while App Engine Flexible Environment expanded to support custom Docker containers and multiple runtimes. In a modernization demonstration, Apigee converted legacy SOAP and WSDL services into RESTful JSON APIs, enabling an event-driven Cloud Function to update an on-premises system from a mobile image upload seamlessly. Intelligent Collaboration and Enterprise Ecosystems (84:38) Prabhakar Raghavan presented machine intelligence as a mechanism to automate repetitive operational tasks, noting that over one in eight Gmail replies are already machine-composed via Smart Reply. G Suite enterprise upgrades included general availability for Team Drives and Google Vault for Drive, the acquisition of AppBridge to migrate on-premises file servers, and Drive File Stream. Google also unveiled rebuilt Hangouts Chat with threaded rooms, the @Meet scheduling bot, Hangouts Meet video conferencing, Jamboard digital whiteboards, and native Gmail add-ons shown in partnership with Intuit QuickBooks. Addressing strategic transformation, Chet Kapoor and Accenture's Gene Reznik emphasized that enterprises must engage across internal, partner, industry, and public digital ecosystems. Executive panelists from Veritas, Pitney Bowes, and Magazine Luiza noted that moving to cloud and API platforms enabled radical business shifts—such as Magazine Luiza increasing digital sales by 35% during an economic recession—while cautioning that modernizing culture and auditing data cleanliness remain critical prerequisites before migration. Google Cloud framed its hyperscale infrastructure, managed data systems, and API architectures as an integrated operating foundation to protect enterprises from competitive disruption. However, executive panelists emphasized that adopting these tools requires deep internal changes, warning that organizations must first overcome organizational inertia and clean up legacy data before realizing the benefits of cloud transformation.

在 Google Cloud Next '17 第二日主题演讲中，Google Cloud 团队及企业合作伙伴阐述了其以底层自研基础设施、弹性数据与算力服务、全栈安全体系、现代化协同套件及 API 生态为核心的企业级云演进路径。高管们主张通过消除传统 IT 的容量规划负担、简化混合云与遗留架构迁移，以及释放员工日常行政琐事，帮助各类企业应对数字化竞争与业务重塑。 自研全球基础设施与弹性算力定价 (0:03) Urs Hölzle 指出，Google 构建超大规模云的基础设施完全从底层技术自研推进，涵盖节能数据中心、定制服务器、专用网络设备、软件定义全球骨干网以及机器学习专用 ASIC。Google 已从多年前投资连接美日 Unity 海底光缆，扩展到在全球铺设高容量海底光纤网，并计划在 2017 至 2018 年间将 GCP 拓展至包含荷兰、加拿大和加州在内的 17 个区域与 50 个可用区。在算力配置方面，GCP 虚拟机即刻支持最高 64 核 vCPU 与 416 GB 内存，并提供抢占式 VM 选项，后续还将推出超过 1 TB 内存的超大实例；同时支持将 GPU 灵活挂载到任意 VM 配置上，按分钟细粒度计费。 针对传统公有云绑定三年预付款导致的资源浪费，Hölzle 援引行业研究称企业平均有 45% 的云支出浪费在未使用的预留容量上，症结在于死板的多年度承诺、固定的 2 次方规格配置以及粗粒度的小时计费。对此，GCP 推出无预付款的灵活 Committed Use 折扣，客户仅需承诺总体核心与内存资源量，便可按月获得最高达 57% 的折扣，且允许动态调整 VM 规格与实例类型；超出承诺部分亦可自动享受持续使用折扣，彻底重构了容量规划与成本控制模型。 分布式数据处理与自动化分析管道 (7:51) 在数据管理方面，Cloud Spanner 展示了分布式数据库范式的突破。Greg DeMichillie 与现场团队演示表明，Spanner 在保留传统 SQL 语义与完整 ACID 事务支持的同时，能够在后台实现节点水平横向扩展与无停机的在线模式更新。在对运行于美、欧、亚三大区域、承载 80 TB 数据量的多区域集群测试中，Spanner 在处理每分钟 50 万笔票务交易时，数据摄取速率达到平均 800 MB/s、峰值约 1 GB/s，无需工程团队手动干预分片、故障转移或数据同步。 在数据分析与流处理领域，Brian Stevens 介绍了源于内部 Dremel 引擎的 BigQuery，以及用于从 AdWords、DoubleClick 与 YouTube Analytics 自动导入数据的 BigQuery Data Transfer Service。针对数据科学家耗费高达 75% 时间清理异构数据的痛点，Google 推出可视化数据准备工具 Cloud Dataprep，通过交互式直方图快速排查异常与数据清洗。纽约出租车车队实时广告推荐演示印证了该闭环：数据经 Cloud Pub/Sub 以超 15,000 事件/秒的吞吐输入，由 Cloud Dataflow 进行流式下采样并送入 BigQuery 归档，同时调用 Cloud ML Engine 模型，实现毫秒级的实时定向广告决策。 应用层零信任安全与敏感数据防护 (33:10) 面对凭证欺诈成为企业首要威胁的现实，Google 强调将安全重心由传统的硬边界与 VPN 转向应用层访问控制。Urs Hölzle 介绍了配合 G Suite Enterprise 实施的物理硬件安全密钥体系，能够阻断基于钓鱼手段的凭据劫持。同时，GCP 推出基于全局负载均衡器的 Identity-Aware Proxy（IAP），管理员无需配置复杂 VPN 或特定客户端软件，即可实现基于身份与安全密钥的细粒度应用级访问；现场通过对部署在 GCP 上的 Oracle E-Business Suite 演示证明，未授权账号被直接拦截，合规企业员工则无需 VPN 即可安全直达内部系统。 在数据安全与合规治理方面，Google 推出 Cloud Data Loss Prevention (DLP) API 与全面公测的 Cloud KMS 加密密钥管理。DLP API 继承了 Gmail 和 Google Drive 的检测能力，支持识别并遮蔽超过 40 种敏感信息类型。在客服支持实测演示中，用户在聊天框与图片附件中过度泄露的社保号（SSN）与电话号码，均在对话结束时被系统毫秒级精准识别并脱敏，包括对实体社保卡照片内部数字进行智能图像掩码，杜绝企业日志分析过程中的合规泄漏风险。 遗留系统上云与无服务器敏捷架构 (43:47) 针对企业在混合云与迁移阶段面临的摩擦，Brian Stevens 阐明了全面搬迁、原生开发与分析平台引入三种主要采用模式。GCP 控制台新增了与底层虚拟化环境无关的在线 VM 实时迁移功能，能够直接将运行于裸金属、Hyper-V、VMware ESX、KVM 或其他公有云的实例无缝转入 GCP。英国零售商 Lush 首席数字官 Jack Constantine 分享了其在高压环境下，仅用 22 天便在圣诞交易旺季前将涵盖 17 个全球电商站点及订单交易的整套生产环境迁入 GCP 的实战案例。 对于现存的老旧接口与松耦合架构改造，Google 通过收购 Apigee 与推进无服务器产品线提供支撑。现场模拟理赔案例演示了移动端拍照上传图片至 Google Cloud Storage 后，触发 Cloud Functions 运行事件逻辑，再经由 Apigee 将底层老旧的 SOAP/WSDL 协议无缝转换为符合现代移动端规范的 JSON 接口并加入流控与认证，随后借助后台迁移向导将原本地部署的服务热迁移入云。此外，Google 宣布 App Engine 推出支持 Docker 容器的灵活环境（Flexible Environment），Firebase 亦在项目突破百万个的节点上强化了与 BigQuery、Cloud Storage 及 Cloud Functions 的全栈连通。 智能协同办公与数字化生态体系 (83:29) 在日常办公协同层面，Prabhakar Raghavan 指出机器智能的职责在于承担繁琐的日常琐事，从而让员工聚焦于高阶创造力。G Suite 推出包含集中化文件权限管控的 Team Drives、合规归档工具 Google Vault for Drive、云端直接流式调用的 Drive File Stream，以及通过机器学习将找文件时间大幅缩短的 Quick Access 功能；同时收购 AppBridge 以加速本地文件服务器迁移。Google 亦重新构建了企业通讯工具 Hangouts Chat 与低门槛视讯产品 Hangouts Meet，并推出支持物理与云端无缝同步的协作白板硬件 Jamboard，且新增 Gmail 原生插件支持以打通 Intuit QuickBooks 等第三方工作流。 在商业模式拓展方面，Apigee 前 CEO Chet Kapoor 与埃森哲高级董事总经理 Gene Reznik 强调，企业必须依托内部、合作伙伴、行业与公共四类生态实现数字化重塑。Kapoor 警示企业迟早都将面临数字化颠覆，必须建立打通本地与多云的互联商业平台。来自 Veritas、Pitney Bowes 和巴西零售商 Magazine Luiza 的高管在座谈中指出，在宏观经济逆风或多云现状下，关键在于尽早识别并淘汰沉淀数据垃圾、利用 API 解耦核心系统，以及坚决推动组织文化从内部流程转向以客户问题为核心的敏捷执行。 Urs Hölzle 总结指出，单纯的技术革新并不足以自动驱动业务转型，还需要与 Pivotal、SAP、Rackspace 等软件及集成伙伴构筑稳固的企业级生态链。尽管演示全方位展示了自底层光缆直至上层 SaaS 的技术敏捷性，但演讲亦反映出：大型企业在跨系统多云资产治理、陈旧代码逻辑解耦，以及突破庞大官僚组织在文化与认知层面的阻力，仍是实现全面云端重塑必须跨越的实际门槛。

[82:58] Prabhakar Raghavan: Machines are really, really good at taking care of the humdrum down there, but it takes humans to do the truly creative work. [42:26] Urs Hölzle: The user's context determines access, not just the network they're on or who they are. [62:06] Brian Stevens: The core promise to developers still remains the same - bring your code and Google will handle everything else. [115:53] Chet Kapoor: APIs are well understood and everybody knows that it's the cornerstone of every digital transformation happening in any and every industry. [23:29] Urs Hölzle: Because we believe that when you move to the Cloud, capacity planning and cost planning should really become a distant memory and not your number one headache. [134:00] Lynn Lucas: The ability to move forward with the people and culture makes all the technology come to life.

[0:03] >> URS HOLZLE: Good morning, everyone. [0:06] Good morning and welcome to the second day of Next. [0:11] So, as you know, we've been building and operating a [0:13] hyperscale Cloud for a very long time, and our goal throughout [0:18] this time actually has remained the same, namely to make it [0:22] possible for you to build great applications and to be very [0:25] productive as a developer or as IT while we deliver this on a [0:31] global scale with unprecedented reliability, [0:35] performance, and security. [0:38] And this Cloud powers all of Google's services, including the [0:42] seven billion user services, so YouTube, Gmail, Search, and now [0:49] it also powers a GCP, which itself is a billion user [0:54] application, meaning that every single day the users of GCP, the [1:00] customers of GCP, connect to over a billion individual users [1:05] every single day. [1:07] And what makes this possible is that we started from [1:11] first principles. [1:13] We designed every element of our infrastructure so that you could [1:18] be uniquely productive and that you can enjoy the performance [1:23] that we created. [1:26] Now when designing this from first principles, you have to go [1:31] and actually optimize every single element, so from [1:35] efficient data centers to custom servers to custom networking [1:39] gear to a software-defined global background to specialized [1:44] ASICs for machine learning. [1:47] We've been living this Cloud at scale. [1:50] And, in fact, in just the last three years, we spent almost $30 [1:54] billion on capital expenditures alone. [1:59] Now one example where this all shows is our network. [2:04] It's probably the largest global network today. [2:07] Analysts put its traffic at between 25 and 40% of global [2:13] Internet user traffic. [2:16] And as a GCP or G Suite customer, you benefit from this [2:20] network because your traffic travels on our private [2:26] ultra-high-speed backbone for minimum latency. [2:29] And 98% of the time we hand off your packets directly to the [2:35] end-user ISP because we interconnect directly with [2:40] almost any ISP everywhere in the world. [2:44] So you won't see any congestion. [2:45] And fewer handoffs mean more throughput, lower latency, and [2:50] better security. [2:53] So, in fact, we have a global network presence in 182 [2:58] countries or territories in the world. [3:01] For comparison, the United Nations today has 193 [3:05] Member States. [3:08] And so to carry this traffic to pretty much everywhere in the [3:10] world, we also need to cross oceans. [3:14] And so nine years ago Google became the first non-telecom [3:18] company to build an undersea cable. [3:21] That was the Unity cable from the U.S. [3:23] to Japan. [3:24] And since then, we've built or acquired fiber capacity, [3:28] submarine fiber capacity, pretty much anywhere in the world so we [3:31] have a redundant backbone to pretty much any place. [3:35] So, for example, last year we turned up what today is the [3:39] highest capacity submarine cable in the world from the U.S. [3:43] to Asia. [3:44] And in just a few months, we'll turn up an even higher capacity [3:48] cable between the U.S. [3:50] and South America. [3:53] Now one of the ways we use this network, of course, is to [3:57] connect GCP regions to each other. [4:00] Last year we announced that we're building 11 new [4:03] Cloud regions. [4:05] Of these 11, Oregon and Tokyo are already live, Singapore and [4:09] Norther Virginia will go live in just the next few months, and [4:12] Sydney and London will come on shortly after that. [4:16] And today I'm excited to announce that we are adding [4:18] three more regions to the set, namely the Netherlands, Canada, [4:22] and California. [4:30] And all of these will come online either this year or next [4:33] and bring the total count of GCP regions globally to 17 and the [4:38] total number of zones to 50. [4:41] Now our technology leadership extends from [4:45] hardware to software. [4:47] If you think about Containers, MapReduce, NoSQL, TensorFlow, [4:51] Serverless, Kubernetes, they all originated at Google and they've [4:56] been powering our services, including GCP, for a long time. [5:01] And recently we added yet another of these ground-breaking [5:06] new systems to our lineup so you can use it on GCP, and I'm [5:10] talking about Cloud Spanner. [5:12] Spanner solves a long-standing problem in databases. [5:16] Until now, you could either use a database, a SQL database with [5:22] nice transactional semantics, but it was really hard to scale [5:26] it beyond one or just a few machines and very complicated [5:30] and very expensive. [5:32] Or you could use a no SQL system like Bigtable and you get [5:36] infinite scale, horizontal scalability, but you no longer [5:40] have strong consistency guarantees, and so you're [5:43] pushing the complexity into your application. [5:46] Neither of these two choices gave you a truly global system [5:50] because your replicas needed to be pretty close to each other in [5:54] order to get good performance. [5:57] And Cloud Spanner, which is a globally distributed database [6:00] service, solves these problems. [6:03] So it scales horizontally like a NoSQL system and thus it can [6:07] take pretty much any transaction loads, but it also has these [6:12] strong consistency guarantees and the relational SQL semantics [6:18] of traditional databases. [6:20] And as an extra bonus, it can also be globally distributed [6:25] while managing high performance. [6:28] We use it for hundreds of mission critical applications [6:32] inside of Google, including AdWords, and now it's available [6:35] on GCP as well. [6:38] So how does Spanner do that? [6:40] I don't really have time to go into that but I'll show you [6:42] one example. [6:44] So Spanner uses a timestamp service, a highly accurate [6:49] timestamp service, to keep the copies of your data in sync. [6:56] This timestamp service is powered by atomic clocks because [7:00] it needs to be really very highly accurate. [7:02] So that means we literally run atomic clocks in every one of [7:06] our data centers to make Spanner possible. [7:10] But with GCP, since Spanner is a service, you see none of [7:15] that complexity. [7:16] In fact, Cloud Spanner is fully managed and serverless so that [7:21] you don't have to worry about machine types, capacity [7:24] planning, replication, software updates, really anything. [7:27] And that's what we try to provide with all GCP services, [7:32] the as much value as possible with as little administration [7:37] as possible. [7:39] And to show you how Cloud Spanner really works, I am [7:42] pleased to introduce Greg DeMichillie. [7:51] >> GREG DEMICHILLIE: Thanks, Urs. [7:53] Thanks, everyone. [7:54] I hope you like live demos because this is the first of [7:57] four live GCP demos you're going to see in the keynote today. [8:01] And I couldn't be more excited to have Spanner be the [8:03] first one. [8:05] To give you a picture of how amazing Spanner is, let's start [8:08] at the beginning, just creating the database. [8:11] Those of you who have actually tried to deploy a multi-region [8:14] database that's highly scalable, highly reliable, and highly [8:17] reliant, know that there's like a zillion steps. [8:20] You've got to configure machines, you've got to set up [8:22] replication, you've got to worry about failover, you've got [8:24] sharding, what happens if I have network issues, and then there [8:28] is the operational side of just keeping the thing running, [8:30] having to have whole staffs of people who do nothing but take [8:33] care of machinery. [8:34] With Cloud Spanner, we turned that into one machine. [8:37] Robert is going to bring up the dialogue to do that. [8:40] He names the instance, he chooses whether he wants it to [8:44] be a single or a multi-regional database, and then he says how [8:47] many nodes. [8:48] And in this case, he picks 42 because that's always the [8:51] right answer. [8:53] So that's it. [8:54] We've now created a multi-regional database that has [8:57] amazing capabilities. [8:59] But that's getting started. [9:01] Let's look at Spanner in an actual real-world scale example. [9:05] Imagine you're a ticket seller. [9:07] You want to sell tickets to events worldwide to [9:10] customers worldwide. [9:11] Now obviously it's pretty important that you sell a ticket [9:14] once and only once, right? [9:16] And it's also pretty important that customers get a good [9:19] performance experience wherever in the world they are buying [9:22] their tickets. [9:24] In this example, we've deployed Cloud Spanner into three Google [9:27] regions - U.S., Europe, and Asia - all connected by that private [9:30] network that Urs told you about. [9:33] We're also running workloads that emulate customers buying [9:36] billions of tickets. [9:38] In this case, we're selling at about half a million tickets per [9:41] minute through this system, all through a system that is [9:44] maintaining consistent version snapshots, full asset support [9:48] for transactions, full, strong consistency. [9:51] That's the power of Spanner as a distributed base, as a [9:54] distributed database. [9:55] Our customers get a low latency experience; we get good old [9:59] traditional SQL consistency. [10:01] Now to prove that this really is a database of some size, let me [10:04] show you the schema behind this database. [10:08] You can see it actually is a SQL database. [10:10] There are columns that are joined together. [10:12] And if you look at that ticket table, it has three billion rows [10:15] in it. [10:17] So while we're doing this, in fact, we're running all this [10:20] traffic against a real SQL database. [10:22] In fact, we can actually run SQL queries against that database [10:26] while it's being hammered with all that traffic. [10:29] Here we have a query. [10:30] And if you look at the bottom part of the where clause, you [10:32] will see that we are looking for tickets in the U.S. [10:36] that are still available as of 10:00 PM and how many seats [10:40] they have. [10:41] So these are our top five events with unsold tickets. [10:45] Now what's important here is what you don't see. [10:46] What's remarkable is how unremarkable this SQL looks. [10:50] If you have deployed a sharding system, your SQL is a mess [10:53] because you're trying to have to do joins across multiple shards [10:56] of your database. [10:57] In fact, it's probably not even human writable SQL when you use [11:00] those systems. [11:01] Spanner takes care of all that. [11:04] But you know real-world applications aren't static; they [11:06] change over time. [11:08] What happens when your needs change? [11:10] Well, first of all, Spanner lets you make transactionally [11:12] consistent schema updates with no downtime, which is critically [11:17] important when you're running systems that your business [11:19] depends on. [11:20] And what happens if you get a wonderful success and your [11:23] needs grow? [11:24] Well, with a traditional SQL database, you better put a PO in [11:27] for a brand new server, you better figure out how to [11:30] configure the new server, you better figure out how to migrate [11:32] the data, you better figure out how to do cutover. [11:35] Or if I'm doing scale out, well now I've got to add a whole new [11:39] complex middleware layer that's incredibly difficult [11:41] to maintain. [11:42] Even that takes time and adds operational risk. [11:45] In Cloud Spanner, Robert is going to go back to the console [11:48] and show you how we add capacity to a Spanner database. [11:52] We simply go from 70 nodes to 99 nodes. [11:55] That's it. [11:56] We've scaled out. [11:57] That's the power of a horizontally scalable [11:59] SQL database. [12:07] And, of course, since we use Spanner so much internally, we [12:11] made sure it handles large database. [12:12] Robert is going to switch over to the monitoring console here [12:15] and what I want you to look at is that graph on the [12:17] bottom left. [12:19] We've been running this whole demo against an 80 [12:21] terabyte database. [12:23] And if you look at the top right graph, when we loaded the data [12:26] in, we loaded the data at an average of 800 megabytes per [12:30] second, and we peaked at about a gig per second of loading data. [12:35] That's amazing. [12:36] That's crazy fast. [12:38] So that's Cloud Spanner. [12:40] You create a highly reliable performant, fully managed, [12:43] multi-region database in seconds. [12:46] Cloud Spanner is SQL. [12:48] You get interoperability in the power of SQL queries, you get [12:51] ASIC transactions, plus you get advanced features like on the [12:55] fly schema changes. [12:57] Cloud Spanner is fully managed. [12:58] We never had to worry about setting up replication, [13:00] sharding, failover. [13:02] The system actually even tunes itself automatically over time. [13:05] So, as you use it, it learns your query patterns and it [13:08] optimizes its own performance characteristics. [13:12] And Cloud Spanner is built for scale. [13:13] We've been running our largest production systems on this [13:16] for years. [13:18] So with that, I'll turn it back over to Urs for a little bit [13:20] more of the keynote. Thanks. [13:28] >> URS HOLZLE: Thank you, Greg. [13:30] So if you're used to traditional databases, it takes a while to [13:33] absorb that. [13:34] I just want to summarize because Spanner is truly something new. [13:40] It is a SQL database, a traditional SQL database, with [13:46] the same semantics, that just scales by changing a value on [13:51] the admin interface. [13:52] It scales through thousands of nodes and has the same strong [13:56] consistency guarantees that your applications have come to [13:59] depend on. [14:00] And, of course, as you saw, easy to administer, does replication [14:04] automatically, multiple locations worldwide, all without [14:08] any setup. [14:09] It has been immensely popular inside of Google, and so we're [14:12] very, very excited that we have it now available on GCP. [14:16] Now from databases to compute. [14:20] Many of our customers have demanding workloads, financial [14:23] risk models, movie rendering, scientific computing, large ARP [14:28] systems, and so on. [14:29] And so starting today, GCP VMs will come with up to 64-cores [14:34] and 416 gigabytes of memory. [14:37] And like our other VMs, these VMs are available as [14:40] pre-emptible VMs. [14:41] So if you have some time or flexibility in your computation, [14:45] you can use them for batch workloads and get much [14:48] lower prices. [14:51] We're not stopping here. [14:52] So, later this year, you will see even higher core counts and [14:55] memory sizes of a terabyte or more. [14:59] In fact, when it comes to hardware innovation and to [15:03] optimizing the hardware for the best performance in the Cloud, [15:07] we've been working very closely with industry partners [15:10] throughout the industry to specialize these systems and to [15:15] accelerate the innovation in hardware. [15:19] One of these partners of course is Intel. [15:21] Please welcome RaeJeanne Skillearn from Intel to talk [15:25] about our partnership. RaeJeanne. [15:35] >> RAEJEANNE SKILLEARN: Hello. [15:36] >> URS HOLZLE: Thank you for joining us, RaeJeanne. [15:38] It's great to have you here. [15:39] >> RAEJEANNE SKILLEARN: Yes. Thank you. [15:40] >> URS HOLZLE: We started to work together quite a long time. [15:43] In fact, I don't quite remember even how long ago it was. [15:46] Can you tell us about the history? [15:47] >> RAEJEANNE SKILLEARN: Yeah. [15:48] It actually started in 2003 with our Intel Core 2 processor. [15:52] Google was able to give us in-depth performance analysis [15:55] and real-world benchmarks and enabled us to tune our processor [15:59] for their unique workload. [16:01] Fast forward to 2008, we started working on the Intel Xeon [16:05] processor, the 5600 at the time. [16:07] And since then, we have optimized six generations of [16:11] Xeon processors specifically for Google's environment. [16:15] Now in November of 2016, we took that collaboration and we [16:19] expanded it. [16:21] Your Diane Greene and our Diane Bryant announced a strategic [16:24] alliance between our two companies. [16:27] We are now collaborating on Hybrid Cloud orchestration, [16:30] security, machine and deep learning, and IoT edge to [16:34] Cloud solutions. [16:36] >> URS HOLZLE: That's right. [16:37] We have had six generations of custom processors and that [16:41] required a lot of work. [16:42] We actually learned a lot from that. [16:44] Can you tell us how we worked together on this and how this [16:48] actually works in practice? [16:49] >> RAEJEANNE SKILLEARN: I can. [16:50] And it starts early. [16:51] It starts in the absolute earliest phases of our [16:54] architecture and CPU process development. [16:57] We take the Google feedback every step of the way and we [17:00] incorporate it and we iterate and modify our processors. [17:03] We know that to truly create the best technology that is [17:06] performance optimized, it really relies on the software as well. [17:10] That's why both of our companies are investing in the TensorFlow [17:13] and Kubernetes open source projects. [17:16] Intel is actually going to contribute code to both of those [17:19] projects so we'll dramatically improve the performance of both [17:22] Kubernetes and TensorFlow on our Intel Xeon processors and our [17:26] Intel Xeon Phi processors. [17:29] >> URS HOLZLE: So three weeks ago we launched the newest [17:34] server, Intel Xeon server processor codenamed Skylake on [17:39] GCP, but yet you can't buy this processor anywhere and Intel is [17:43] not planning to launch it for quite a while. [17:46] >> RAEJEANNE SKILLEARN: For a while. [17:48] >> URS HOLZLE: How does that work? [17:49] >> RAEJEANNE SKILLEANER: Well, first I would like to [17:51] congratulate Google on being the absolute first to have Cloud [17:54] services on our next generation Xeon processor, Skylake. [17:58] >> URS HOLZLE: Great. Thank you. Thank you. [18:01] >> RAEJEANNE SKILLEANER: There was a trick to doing this. [18:03] First we had to accelerate production readiness of a [18:06] targeted set of features that are a little bit different than [18:09] our broad feature set, all the systems in the software that we [18:12] have to qualify and validate for the broad general availability. [18:16] Intel and Google started an early definition on a custom [18:18] skew all the way through significant joint investment in [18:22] the validation of those systems into production in [18:25] your environment. [18:26] It was a team effort many steps of the way, at massive scale and [18:31] in a very aggressive timeline for both of us. [18:34] >> URS HOLZLE: Well thank you for being such a great partner. [18:35] We really enjoyed working with you. [18:37] >> RAEJEANNE SKILLEARN: Thank you. [18:37] We enjoy it. [18:38] We look forward to tomorrow. [18:42] >> URS HOLZLE: We're very happy that our customers benefit from [18:45] this decade of collaboration between us and Intel. [18:48] Because we're pushing the envelope in so many directions [18:52] on performance that we have to really, really work very [18:57] differently with vendors. [18:59] Skylake offers great performance for compute-intensive workloads. [19:03] We are very, very happy that Skylake is available at first on [19:07] Google Cloud. [19:09] Now let's go from performance to economics. [19:13] Clouds, as you know, are supposed to be elastic and free [19:16] from capex or capacity planning, yet some Cloud providers force [19:21] you to pay up front for three years to get the best price. [19:26] But if you have to buy a VM for three years, then how is that [19:30] better than buying your own server for three years? [19:33] There's no flexibility. [19:35] And, in fact, a recent study showed that Cloud users waste on [19:39] average 45% of their spend on resources that they bought but [19:45] can't use. [19:47] That's caused by many factors. [19:50] One is these three year leases that force you to predict your [19:53] future perfectly. [19:54] And let's face it; none of us can do that. [19:57] We end up with stranded resources that we actually [20:00] can't use. [20:01] Similarly, we are forced to buy - like on premise - we are [20:05] forced to buy servers in fixed sizes, and so we end up [20:09] overbuying on some dimension, and we still pay for it even [20:13] though we can't use it. [20:14] And last but not least, you pay for the full hour even when your [20:17] test run runs for only ten minutes. [20:19] So you pay for compute time that you're not even using. [20:22] All of that adds up to 45% waste. [20:27] Now we believe that it should be easy to get the best price, and [20:30] so we have solved these sources of waste. [20:33] In fact, in 2014 we introduced automatic sustained use [20:36] discounts that give you discounts without any [20:39] long-term agreement. [20:40] So as soon as you use a VM for more than a quarter of a month, [20:45] you start seeing savings. [20:46] And if you use the VM for the entire month, you get a [20:49] 30% discount. [20:52] But you're still free to stop using a VM at any time. [20:57] So sustained use discounts brought the Cloud back [21:00] into Cloud. [21:01] You can change your mind at any time but you still get [21:03] great prices. [21:05] And you don't need to accept VMs whose shapes come in powers [21:10] of two. [21:11] So with other providers, if your application needs say 20 cores [21:15] and 50 gigabytes of RAM, you might have to buy the next [21:20] larger machine size that might have 40 cores and 160 gigabytes [21:25] of RAM. [21:26] And so you end up paying literally twice the resources [21:30] that you need. [21:31] But not on GCP, because with custom machine types, you can [21:36] dial in exactly the configuration you need so you [21:39] pay for exactly the resources you need. [21:43] Today, over 20% of our core hours on GCP are for custom [21:48] machine types and our users save an average of 19% through [21:52] the customization. [21:54] And we even help you save money with our right [21:57] sizing recommendations. [21:58] That's a service that looks at your memory and CPU usage and [22:02] then suggests the best virtual machine size. [22:06] And last but not least, GCP has permanent billing for VMs. [22:09] So if you use a VM for 11 minutes, you pay for 11 minutes. [22:15] It seems pretty logical. [22:17] So GCP and only GCP is truly an elastic Cloud. [22:23] You only buy what you need, you only pay for it when you [22:28] actually use it, you get automatic discount if you use it [22:31] for extended periods of time, and we even alert you when you [22:34] appear to be wasting resources. [22:36] And when you put all of this together, you save an average of [22:40] 60% relative to what you would pay on other Clouds. [22:46] So given the complexity of Cloud pricing we just went through, [22:49] it's not surprising that the same study I quoted says that [22:53] 53% of Cloud users say that optimizing and controlling spend [22:59] is their top problem in the Cloud. [23:02] But not with GCP. [23:04] Because with GCP, you don't need to create an entire new ministry [23:08] in your company just to get the best price because our flexible [23:12] pricing structure lets you enjoy a Cloud as it was meant to be - [23:16] on demand, pay as you go, paying only for what you need. [23:22] But today it gets even better because we're introducing [23:25] another way for you to save - committed use discounts. [23:29] So in exchange for a one or three-year commitment, you [23:32] receive a discount of up to 57% billed monthly, [23:36] no upfront payment. [23:37] But these are not inflexible reserved instances that lock you [23:41] into a particular instance type or family that force you to pay [23:47] up front. No. [23:48] They only commit you to an overall volume for your compute [23:52] and memory. [23:54] So you're not locked into any particular VM size. [23:58] You can change individual machine types and VM shapes [24:01] at will. [24:02] You can change the number of VMs. [24:03] You're only committing to the aggregate volume and not [24:07] the details. [24:09] And if you're not sure how much commitment to make, you can [24:13] start smaller because you still get sustained use discounts and [24:19] all the other benefits on any of the usage above your [24:22] committed usage. [24:23] And so you're not facing a huge unit cost cliff when you exceed [24:29] your forecast. [24:32] Because we believe that when you move to the Cloud, capacity [24:36] planning and cost planning should really become a distant [24:39] memory and not your number one headache. [24:42] So on GCP you are saving money automatically with no regrets, [24:46] no spreadsheets, and no PhD in economics needed to manage [24:50] it all. [24:52] In fact, one way where you see this flexibility is our GPUs. [24:57] Many of our customers already GPUs for a variety of [25:00] applications to speed up simulations, transcoding, deep [25:04] learning, computational chemistry, finance, and [25:07] many more. [25:08] But on GCP, you can add a GPU to any VM configuration and start [25:14] it up in under a minute, much faster than our competitors. [25:17] So that means that if you want a GPU with lots of memory, you can [25:21] get that. [25:22] Or if you want a GPU with little memory but lots of cores and an [25:27] SSD, you can get that, and you don't waste money on your GPU [25:32] because you have to buy some other stuff with it. [25:36] And, of course, GPUs also have per minute billing so you don't [25:40] end up paying when you're not actually using it. [25:44] So to help you learn more about this, I want to introduce one of [25:47] our customers who is using GPUs as a critical part of [25:52] their business. [25:53] Please welcome Ashok Belani from Schlumberger. [26:06] Hi, Ashok. [26:07] Great to have you here. [26:08] >> ASHOK BELANI: Thank you for having me. [26:10] >> URS HOLZLE: So Schlumberger moved to Google platform [26:11] last year. [26:13] What drove you to adopt Google? [26:17] >> ASHOK BELANI: At Schlumberger, we've been leaders [26:19] in high performance computing for many decades. [26:23] You see the size of our compute clusters growing on the slide. [26:28] We bought the first grade back I think in the mid '80s and we [26:32] moved to massively powered PCs in the '90s. [26:36] And then we were actually one of the first companies to jump onto [26:40] very high GPU to CPU ratios back in 2007 or 2008. [26:47] What we think is that we can go to the next level of compute [26:50] possibilities working together with Google. [26:54] >> URS HOLZLE: So you clearly use massive amounts of compute. [26:57] Tell us what you're actually doing with compute [27:00] and those GPUs. [27:01] >> ASHOK BELANI: So we are looking in the subsurface for [27:04] our geologists who are exploring for oil and gas deposits. [27:09] So we look very deep into the subsurface, tens of thousands [27:12] of feet. [27:13] We use acoustic data and dynamic data of different kinds. [27:18] Here you see on the slide behind me a vessel which has a huge [27:23] spread of sensors which are being pulled behind it. [27:26] There are millions of sensors on a spread which is ten kilometers [27:30] by two kilometers. [27:32] Sometimes we use two of these boats which would cover half the [27:37] surface area of San Francisco. [27:38] And, as you see, the amount of data that we generate has been [27:41] increasing over the years significantly. [27:43] And then we use high performance computing to create images of [27:47] the subsurface for our exploration efforts. [27:50] >> URS HOLZLE: So tell us how your experience has been. [27:52] How has it been working with Google? [27:54] And tell us actually what we're seeing here. [27:56] >> ASHOK BELANI: So we see an image here which actually has [28:01] been - the data has been acquired over about five months. [28:05] This data was actually processed in Google over a period of [28:09] three months. [28:10] And you see the subsurface which is actually 20,000 feet into the [28:14] ground and it is 50,000 square kilometers off the shore of [28:19] Mexico - bigger than the size of Switzerland. [28:25] >> URS HOLZLE: Our mountains are taller. [28:28] No need to go on the ground. [28:32] >> ASHOK BELANI: I think the interesting thing about working [28:34] with the Google Cloud is that we are able to create basically a [28:38] cluster which is suited for that particular problem because of [28:42] the algorithms we are going to use on that problem. [28:44] Within a matter of minutes, we are able to mix the CPUs, GPUs, [28:47] high GPU ratios, memory close to the processor the way we need [28:52] for particular algorithms to work in a very good way. [28:55] This is a big advantage that we get out of Google. [29:00] But I would say another advantage is Urs Holzle himself. [29:02] I think Google should be very proud of having someone [29:04] like this. [29:07] >> URS HOLZLE: Okay. Yeah. [29:10] Let that be stricken from the record. [29:11] That was not in the script. [29:13] >> ASHOK BELANI: That wasn't in the script. [29:14] We certainly cooperate very well with the engineers in Google. [29:18] We actually created a center here in Menlo Park so that our [29:22] engineers could work together with Google very closely. [29:24] And we are only at say 10-15% of the way on this journey to make [29:32] Cloud very efficient on our applications. [29:38] We think we will be able to weave in things like big data [29:42] type of technologies or analytics or machine learning. [29:47] In the future, they will be woven into our applications and [29:50] I think we will definitely be able to achieve the next level [29:52] of computing in oil and gas. [29:54] And I think we serve all the oil companies in the world. [29:58] So as we go on this journey, they will come together with us. [30:03] >> URS HOLZLE: Thank you very much. [30:04] It's really been great working with you. [30:05] >> ASHOK BELANI: Thank you very much. [30:05] >> URS HOLZLE: Thank you, Ashok. [30:11] All right. [30:12] So when you use Google, you use the same security infrastructure [30:18] that Google uses. [30:20] And we've been investing an enormous amount of effort into [30:23] that security. [30:25] You can see an overview of the many layers here. [30:27] I don't really have time to talk about that today. [30:29] Today I would like to show you just a few highlights. [30:32] But we have a detailed security design, white paper, and we have [30:36] a one-hour session that do real justice to this topic. [30:42] So we put a lot of effort into security starting with [30:46] physical security. [30:47] For example, in this single Google data center campus we [30:50] have over 175 security guards on staff 24/7 in addition to [30:56] countless cameras, motion sensors, iris scanners, and [31:00] so on. [31:02] Many of our locations are not individual data centers but data [31:06] center campuses. [31:07] And so this very high level of physical security is amortized [31:12] over hundreds of thousands of servers. [31:15] And thus on GCP, low prices don't mean low [31:20] security standards. [31:22] In fact, to protect the security of our hardware, we put a [31:28] security chip on all our new machines to serve as the basis [31:32] of trust for that machine's identity. [31:35] So this is a custom chip designed by Google so we know [31:38] exactly what it does and it helps us protect servers against [31:43] tampering even at the bios level; or in this particular [31:46] example, it helps protect the bios of the networking card that [31:52] we built. [31:54] The chip, in fact, is so small that I'm actually wearing one on [31:59] my earring here. [32:02] So here it is. [32:06] So now you know that you are watching the authentic [32:14] GCP keynote. [32:21] So this hardware chip helps us authenticate the hardware. [32:25] And then on top of that, it helps us authenticate the [32:28] services that we run. [32:29] When services call each other, so the services that implement [32:33] GCP, they must mutually prove their identity to each other and [32:38] use certificates for that and the binaries are signed and [32:42] cryptographically signed so that we can verify we are running the [32:46] right binary. [32:48] And then on the storage side, GCP's durable storage services [32:52] encrypt all data before it is written to the physical media. [32:57] And on the networking side, all Internet traffic from and to G [33:02] Suite or GCP are protected with strong encryptions and multiple [33:06] layers of service protection. [33:10] Now that covers the data center side. [33:13] But a system is only as secure as its user accounts. [33:18] Today, phishing probably is the number one security problem for [33:22] enterprises, meaning someone is trying to trick your users into [33:26] providing their password and perhaps their one-time token. [33:33] But your G Suite and GCP account is already protected by a [33:36] sophisticated abuse detection system to thwart those attackers [33:40] that are trying to guess your password or trying to use a [33:43] stolen password. [33:44] But with our optional phishing-resistant second [33:48] factor, we can provide a very strong defense against phishing. [33:54] No other Cloud today provides you this protection against what [33:58] is probably the number one security problem in enterprises. [34:03] For further end-to-end security, we also ensure that the user's [34:07] client device is secure. [34:08] So Chrome, Chromulus, and Android are designed ground-up [34:14] for security, all featured Cloud management and frequent [34:17] over-the-air security updates. [34:19] And our client operating systems feature a hardened boot. [34:22] So they actually have a chip similar to this to verify that [34:27] they're booting the correct software. [34:29] They have encryption of storage by default and enterprise [34:33] grade management. [34:35] And with such a strong security stack and nearly zero [34:39] administration cost and a wide range of models to choose from, [34:43] it's no surprise that Chrome Books outsold Macs last year. [34:50] >> URS HOLZLE: Yes, thank you. [34:52] So today we are announcing several new security features [34:55] that we're adding to GCP and G Suite. [34:57] First, some tools to let you secure your data. [35:01] Data loss prevention is an API now available that lets you [35:05] discover PII and other sensitive data in your content and take [35:09] appropriate actions such as redacting it. [35:12] You will see a demo in a minute. [35:14] The engine behind this API is the same engine that powers the [35:18] data loss prevention feature in Gmail and Drive, so you get [35:21] consistent results everywhere. [35:24] Next, Cloud Key Management Service is now [35:27] generally available. [35:28] It lets you manage encryption for your Cloud services because [35:32] Cloud KMS protects your data and secrets that are stored at rest [35:37] and it can automatically rotate your keys for you. [35:44] And then we also added tools that let you safely access that [35:48] secured content once it's been secured. [35:50] The first one is the Identity Aware Proxy or IAP. [35:54] It is available in beta and it enables you to configure secure [35:59] controlled access to your applications. [36:01] So you can enforce a who can see what access control at the [36:06] application layer. [36:08] So you don't need client software, remote access VPNs, [36:12] firewalls, network configurations. [36:14] You just deploy IAP with a single click and you're going to [36:17] see that in a second. [36:18] So IAP acts as a smart front-end in front of all of [36:23] your Cloud applications. [36:25] It's built on top of the GCP load balancer so you benefit [36:29] from the transport security and the scalability of the Google [36:33] front-end service. [36:35] And last but not least, security key enforcement is now available [36:40] to GCP users as part of G Suite Enterprise. [36:44] So this feature lets you enforce security key use for all members [36:50] of your domain. [36:52] That means that all of your applications - access to all of [36:55] your applications is now strongly phishing-resistant. [36:59] Here to demonstrate some of these features are Greg [37:03] DeMichillie and team. [37:10] >> GREG DEMICHILLIE: Thanks again, Urs. [37:12] You know, today the task of taking a corporate application [37:16] and making it available to your employees outside of your [37:19] network is just painful. [37:21] You've got to set up VPNs and nobody likes those. [37:24] They're hard to set up. [37:25] They're hard to configure. [37:27] And in the end, you end up with this perimeter-based security, [37:29] right, with a semi-hard exterior and a soft underbelly, which [37:33] isn't nearly as effective as modern device [37:35] multi-factor authentication. [37:38] So we're going to give you an example of how easy it is to use [37:40] the Identity Aware Proxy to take a typical corporate enterprise [37:44] app and make it available to your users. [37:46] Now the specific app we're going to use in this case is as [37:48] enterprise as it comes. [37:49] It's Oracle E-Business Suite running on Google [37:52] Cloud platform. [37:53] And as a company, we want our users to access this application [37:57] whether they are in the office or they are out on the road. [37:59] So let's see how simple that is to do with IAP. [38:02] Neil is going to click the button to turn on IAP. [38:05] He's then going to tell us what domain he wants to publish that [38:08] application for. [38:10] And now users in that domain and only that domain can access [38:14] the application. [38:15] So let's test it out. [38:16] He's going to flip over to a browser tab. [38:19] Now the first thing is there is no VPN software installed on [38:22] this machine. [38:22] We are just navigating to the URL. [38:25] Now first he's going to try his personal Gmail account which has [38:27] not been authorized for this application. [38:30] Sure enough, he is denied. [38:32] Now he's going to try it with his corporate account which has [38:36] been authorized and he gets the username, he gets the password, [38:42] but now he will get prompted to enter his second [38:45] key authentication. [38:46] So if you look on the sides, I think we're going to project [38:49] that up in a second. [38:52] Well, okay, trust me, he pushed - there it is. [38:54] He pushed the security key. [38:56] I think you know what that looks like. [38:58] And now he's taken directly into the application without having [39:02] to have another additional login step needed. [39:05] Not only is this easier for the developer, it's easier for the [39:08] admin, and lord knows it's easier for the end user rather [39:10] than hassling with a bunch of VPN software, and it's more [39:13] secure because we're not trusting anybody who just [39:16] happens to come in through a VPN tunnel. [39:18] So that's IAP. [39:21] The other thing Urs talked about was data loss prevention. [39:24] Let me show you how Google can help you secure some of your [39:26] most sensitive data. [39:28] Most companies today have a policy around data minimization, [39:32] the goal being to minimize the amount of data that you collect [39:35] and store to only the data that's actually needed to run [39:38] your business. [39:40] But that's easy to do in words but hard to do in practice. [39:44] Data has a way of leaking out all over the place. [39:46] So let's look at a concrete example. [39:48] We have an enterprise application that a support agent [39:51] might use to chat with your end users. [39:54] And in the chat you can see Neil is a support agent and he's [39:57] asked Alice for some information to verify her account. [40:00] Well, Alice has way over-shared. [40:02] We just wanted the last four of her Social. [40:04] She gave us her full Social, her phone number, and a picture of [40:07] her Social Security card. [40:10] Now obviously we want to store this chat, right, to see how our [40:15] conversations are going or how the agents are working, but I [40:18] can't store that in a database. [40:20] I'm going to have half my company now knowing Alice's [40:22] Social Security Number. [40:23] So how do we do that? [40:25] DLP makes it easy. [40:26] When he ends the chat, we will use Data Loss Prevention to [40:29] identify and redact sensitive information. [40:33] So now you'll see that we've - yes. [40:41] We recognize over 40 different types of sensitive info types. [40:45] In this case, we've identified Alice's name. [40:47] If he scrolls down a little bit, you can see we found her phone [40:50] number, her Social Security Number, we replaced all those [40:53] with red dots, and even we went into the picture of the Social [40:56] Security card, found the Social Security Number there and [41:00] redacted it. [41:01] Now this chat can be used for analysis. [41:07] Now just to prove this is real, Neil has a webcam over there and [41:12] we're going to try one more example. [41:14] I was in the green room a while ago and I found this credit card [41:17] sitting there and it said Urs Holzle. [41:21] I'm just kidding. [41:22] Neil has a sample credit card with a valid number but it's not [41:25] Urs' card. [41:26] I trust DLP. [41:27] I don't trust 10,000 of you with cell phone cameras. [41:31] So he's going to put the credit card number on it, he's going to [41:33] hit the button, we're going to use DLP, and it will find and [41:37] redact the credit card. [41:40] Now this isn't just dumb OCR. [41:45] You'll notice it didn't block out the expiration date or the [41:48] name because we told the API that the only information we [41:51] wanted to eliminate was credit card numbers. [41:54] So you can control your definition of PII, what matters [41:58] to you. [42:00] So there you have it. [42:00] Identity Aware Proxy. [42:02] We made it super simple to take a corporate application and make [42:05] it available to our end users without the hassle of complex [42:07] network configurations. [42:08] And with Data Loss Prevention, we were able to help it make [42:11] sure that you minimize the amount of data that you collect [42:14] so that's one less headache you have in terms of compliance or [42:17] regulatory or really just running your business. [42:19] And with that, Neil and I will turn it back over to Urs again. [42:22] Thanks, everybody. [42:26] >> URS HOLZLE: Thank you. [42:27] Thank you, Greg. [42:29] So, as you just saw, it's really easy to use the IAP Proxy to [42:32] control access to your applications. [42:35] But we're already working on the next version based on a [42:38] principle that we've been applying to our own corporate [42:41] users for a while. [42:43] Because we view every access decision to resource as not [42:49] something that's just about the user credentials and maybe their [42:53] second factor but really about something that should be based [42:57] on the context around it; for example, the state of the user's [43:00] device, their location, and so on. [43:03] So we call this Context Aware Security. [43:06] The user's context determines access, not just the network [43:10] they're on or who they are. [43:13] The context that we have today with the IAP Proxy is just the [43:16] user identity and the security key - so the strength [43:20] of authentication. [43:21] But you can expect our future versions to use a richer and [43:25] richer context over time to better secure access to your [43:29] Cloud applications or your G Suite. [43:33] Now I am thrilled to introduce Brian Stevens to tell you more [43:36] about how customers adopt GCP. Brian. [43:47] >> BRIAN STEVENS: Good morning. [43:50] Thanks, Urs. [43:52] So, public Cloud has absolutely exited the early adopter stage. [43:56] It's now a shared platform, available to everyone from [43:59] start-ups to the world's largest enterprises. [44:02] And it's quickly gone pan vertical from financial [44:05] services, healthcare, industrial, government, [44:09] everybody is choosing the Cloud. [44:12] And when they pick GCP, why do they come to GCP? [44:15] A lot in common. [44:17] They want the world's best security. [44:20] They want to be on a flywheel of continued innovation, all [44:24] available through an API. [44:28] They want the best in the world data and analytics. [44:31] And the most important thing, they want the tools and the [44:34] platforms that their developers love. [44:38] I've spent a lot of time working directly with customers. [44:40] And when you actually zoom out and you try to look at patterns, [44:44] we're actually seeing three distinct things when they come [44:46] to GCP. [44:48] The first is wholesale migrations of their existing [44:50] workloads from on-premise to GCP, which, to be honest, was a [44:54] little bit of a surprise. [44:56] The second is building Cloud native applications. [45:00] That includes start-ups as well as some of the world's largest [45:03] enterprises now. [45:05] And the third is just coming to Cloud to get the richest set of [45:09] data analytics and machine learning. [45:12] So our first Cloud service was back in 2008, Google App Engine, [45:16] almost nine years ago, and it was an incredibly powerful [45:19] platform as a service, possibly too advanced for its time. [45:24] We're going to talk about that a little bit more soon. [45:26] But most enterprises either can't or don't want to rewrite [45:30] their application architecture just to move to [45:33] the public Cloud. [45:35] And what's amazing is that even without that rewrite, the [45:38] drivers for moving to GCP are still incredibly compelling. [45:44] They get the amazing security that you just saw, they get to [45:47] reduce their capex, they get incredible reliability, Google's [45:52] pretty good about reliability, performance, they get it for [45:56] free, look what you just saw with Skylake and what's [45:59] happening in the network, and they get this [46:01] improved efficiency. [46:03] So I actually kind of loathe, to be honest, industry conferences, [46:09] but there's this one CIO conference that I go to every [46:12] year, and it's put on by Excel, and it's a small, little, [46:15] intimate venue with about ten CIOs. [46:17] And two years ago the top things when we went around the room [46:21] that they cared about was security, what kept them awake [46:23] at night, and mobile handsets. [46:26] The two are really related. [46:28] A month ago when we went back again, same group of CIOs, they [46:31] went around the table and the top two things that were keeping [46:34] them up, security still, and then shutting down data centers. [46:38] Every one of them was shrinking their data center footprint. [46:42] I felt a little weird because - and then it was my turn and I [46:44] was talking about we're doing the opposite. [46:47] Like look at what Urs just showed you. [46:49] So it's kind of clear where these workloads are going. [46:53] So, ideally, you want to be able to shift the workloads with as [46:56] little change as possible because you don't want it to be [46:59] a lot of work just to move to Cloud, and the more change you [47:03] introduce, there's more risk. [47:06] But that's just the first chapter. [47:08] Once you get to GCP, often what happens is chapter two, and [47:12] that's when they actually look at how do we re-factor? [47:15] How do we use things like a managed database? [47:17] How do we use something like Spanner? [47:20] So our call to action at Google is how do we make that as simple [47:24] and easy as possible for them, and that includes technology, [47:27] processes, and people. [47:30] One thing that we did is we recently added right from the [47:33] Cloud console the ability to migrate a virtual machine [47:36] to GCP. [47:37] This is more than transcoding an offline image. [47:41] It's actually the live migration of a running server to GCP. [47:46] What I love about it is that it's hypervisor agnostic, so it [47:48] works on bare metal, it works on Hyper-V, ESX, KVM, even from [47:53] another Cloud. [47:55] Also, we've been making big investments in Windows because [47:58] we want to make GCP as great for Windows developers as it has [48:02] been for Linux and Open Source developers. [48:06] Our goal wasn't just to be an okay Windows platform. [48:09] We want to be a great Windows platform, perhaps the best [48:12] Windows platform. [48:14] So we already have support and pre-built images for many [48:18] flavors, a Windows server, as well as SQL server. [48:21] And we also support active directory running in the Cloud, [48:24] and you can integrate that with your on-premise [48:26] domain controller. [48:29] But it's really important for us that for developers and Windows [48:33] developers that we meet them exactly where they are. [48:36] We don't want them to have to change how they do things just [48:38] to be able to take advantage of Cloud. [48:40] That's why we did the Visual Studio integration that we've [48:43] done with GCP so that you can actually deploy dot net apps [48:47] from Visual Studio and then just manage all of your [48:49] Cloud resources. [48:51] And we've also integrated with hundreds of cmdlets for [48:54] PowerShell right into our Cloud SDK. [48:57] And so now they can be very comfortable managing all of [49:01] their Windows-based Cloud projects. [49:04] So today we're announcing the general availability of SQL [49:07] Server Enterprise and that actually includes support for a [49:10] high availability as well as clustering. [49:13] Also, a beta dot net and that will be available in both App [49:17] Engine and Container Engine. [49:19] And to help people on this journey, we're announcing a new [49:22] Windows partner program. [49:23] And so we've partnered up with a number of top specialists that [49:26] actually have great Windows expertise as well as [49:30] GCP expertise. [49:31] So they can help customers on their journey to move Windows [49:34] environments to GCP. [49:37] So now Linux or Windows, my SQL Oracle or SQL server, and as of [49:43] today also a beta of Managed Postgres, developers can really [49:48] easily migrate to GCP with minimal refactoring of their [49:51] application stacks. [49:53] So what's the best test for moving to the Cloud with [49:56] minimal friction? [49:57] Probably to be able to do that with zero downtime. [50:01] It seems pretty impossible, but Evernote actually did it [50:05] last year. [50:06] They moved their entire software infrastructure from their [50:09] on-premise data centers to GCP - 200 million customers depend on [50:16] Evernote every day - and they did this migration in 89 days [50:21] with zero downtime. [50:24] And so Lush, a British cosmetics company, they began their [50:27] migration to GCP. [50:28] They started last year in September and it was critical [50:33] that they finish in time for the holiday season and they did it. [50:39] And so here, I would like to introduce Jack Constantine from [50:42] Lush to tell us about their journey. [50:55] Hi, Jack. [50:56] >> JACK CONSTANTINE: Hi, Brian. [50:57] >> BRIAN STEVENS: So 22 days. [50:59] I don't think anybody is going to believe you. [51:01] It seems impossible. [51:02] How is that even possible? [51:03] >> JACK CONSTANTINE: Yeah. [51:04] So in September we started discussing doing the [51:07] whole migration. [51:08] The actual migration was from December the 1st until December [51:11] the 22nd. [51:12] So it's not just a little bit business critical; we're talking [51:15] peak trade time. [51:17] Yeah, so it was a huge deal. [51:20] I mean one of the key things, I couldn't have done it without [51:23] the in-house engineering team that we have, some of the guys [51:26] in the audience. [51:27] I wouldn't be up here if it wasn't for all the hard work [51:29] those guys put in, the hours, the dedication, the focus. [51:33] I think they deserve a round of applause. [51:40] We found ourselves in a bit of a tricky spot. [51:43] We were in a contract that we weren't really comfortable with. [51:45] We wanted to be able to actually have a look and see what else we [51:47] could go for. [51:50] The contract ran out on the 22nd of December, hence the reason [51:52] that we had that hard deadline. [51:53] >> BRIAN STEVENS: Lucky us. [51:54] >> JACK CONSTANTINE: Yeah. [51:56] I'm a little bit of a risk-taker myself, as you can tell. [52:02] Basically there is very little bureaucracy in Lush, so the [52:06] ability for me to be able to actually make that decision was [52:10] quite fast and then the team just powered through. [52:12] It was a really exciting time. [52:17] We were really, obviously, so pleased with the result. [52:21] >> BRIAN STEVENS: So were the like any sort of challenges with [52:23] the migration, technical inhibitors, process? [52:26] >> JACK CONSTANTINE: Well I think, like with any migration, [52:28] there are always technical challenges. [52:29] You've got to worry about getting your data from one place [52:32] to the other and the amount of it and are you going to make [52:34] sure you've got consistency. [52:36] We moved 17 websites from all over the world with customer [52:40] data, with order data, product data. [52:42] Obviously you want all of that to be completely up and stable. [52:46] But I think one of the main things really from my [52:50] perspective was the kind of commitment to actually [52:53] achieving it. [52:54] I think sometimes people can - obviously when you're going to [52:56] throw a kind of hard deadline like that, it can sound a [52:59] bit unachievable. [53:00] But I always like to think it depends on kind of which reality [53:04] you're looking from. [53:06] I like to think that it's realistic; you may think it's [53:08] not realistic. Right? [53:10] I think it's realistic so I think we should try and do it. [53:12] So keeping that focus on people actually kind of believing that [53:15] this is a goal we can achieve in the timeframe and not letting [53:18] people start to put the blockers up and go, oh, we're going to [53:20] have to delay this, we're going to have to delay that. [53:22] All of a sudden you watch everyone - I mean, yeah, [53:24] everyone gets very tense, but also you achieve a lot. [53:28] You actually get through it and the things you need to get done [53:30] get done. [53:32] >> BRIAN STEVENS: Anything that could have made it a little [53:34] bit easier? [53:35] >> JACK CONSTANTINE: Oh, definitely. [53:37] I think the awkwardness of our previous supplier and the fact [53:41] that it was a very closed environment made it very [53:44] difficult for us to get visibility of everything we [53:47] wanted to be able to move over. [53:48] >> BRIAN STEVENS: Open wins again. [53:49] >> JACK CONSTANTINE: Exactly. [53:50] Openness is something that we absolutely cherish in Lush. [53:54] Obviously you guys do at Google, which has been great for us. [53:57] And the other thing that we had, we had this great - one of my [54:00] colleagues had this great conversation with a Google [54:01] partner on the phone and it was about a week before and they [54:04] were obviously getting a bit scared, oh, is this going to [54:06] happen, and they were on the phone. [54:08] The partner said what is Plan B? [54:10] My colleagues said Plan B is to make Plan A happen. [54:14] >> BRIAN STEVENS: That's a great line. [54:15] >> JACK CONSTANTINE: That's it. That's it. [54:17] And we made it happen. [54:19] >> BRIAN STEVENS: That's cool. [54:22] Urs, the last speaker, was the one who recruited me to Google. [54:26] I was already sold because I believed the future is public [54:28] Cloud and just fascinated by it. [54:31] You have to be a technology company to win this and just [54:33] everything that Google has been investing in for years [54:35] in technology. [54:37] But the surprise for me when I actually got here was the [54:39] culture, right, the ethics, the diversity, the focus on [54:45] inclusion, sustainability. [54:47] And in our chat yesterday, it sounds like there are a few [54:50] similarities in your culture that you and your parents have. [54:53] >> JACK CONSTANTINE: Yeah. Yeah. [54:54] So my parents founded Lush over 20 years ago and there are a lot [54:58] of ethical values that we've built throughout the [55:01] organization, and we pride that in kind of everyone we go [55:04] through, supply chain when we're buying ingredients, and fair [55:07] trade, and looking at the best quality ingredients, we look at [55:10] that with our packaging. [55:12] I represent the more digital side of the business, obviously. [55:16] It's a very interesting time and Lush is looking at its digital [55:20] future and understanding where we go and how we navigate [55:24] through the landscape. [55:25] I think there are a lot of similarities with Google in [55:28] terms of the openness versus closed, the cultural elements. [55:33] It's been great to be able to - one of the reasons we were so [55:36] eager to do the migration in that time period was because it [55:40] felt like by moving over to the Google Cloud platform we would [55:42] be aligning with our ethical values on a much higher level. [55:45] Things like the renewable energy and the open mentality, all of [55:50] those things, we're looking at the moment around digital [55:53] ethics policy. [55:54] Actually, just before we did the migration in November, we did a [55:58] global campaign to support keeping the Internet on, [56:02] especially in countries where the government may shut the [56:05] Internet down because they don't want to encourage communication. [56:09] Obviously kind of the reverse of what we wanted to happen with [56:11] our migration. [56:12] Thankfully we also kept the Internet on when we migrated [56:15] our websites. [56:17] But, yeah, the synergy between Google and Lush is great. [56:21] I'm really excited about even the things you've been showing [56:24] today and being able to work together much more on [56:27] prototyping new ideas, having that flexibility. [56:32] I spoke to Urs earlier and he was saying about that whole kind [56:34] of engineer-to-engineer dynamic. [56:36] And my team straight away, they were absolutely buzzing. [56:39] We've only been working with you guys for the last 3-4 months but [56:42] the energy is huge. [56:44] It's great. [56:45] >> BRIAN STEVENS: Well, my wife and daughter have always been [56:47] big fans, but I'll say I'm a convert now as well. [56:49] Thanks, Jack. [56:50] >> JACK CONSTANTINE: Thanks very much. [56:57] >> BRIAN STEVENS: That's a good story. [56:58] So Evernote and Lush are great examples of a complete migration [57:02] to GCP from either another Cloud or on-premise. [57:05] But for the largest enterprises, the move to Cloud isn't a point [57:10] in time. [57:11] It's going to be a perpetual state to run in a [57:13] hybrid environment. [57:14] And what we don't want is we don't want this steel curtain [57:17] between public Cloud and private. [57:19] It should really feel like this data center extension, albeit [57:23] this really amazing data center extension. [57:26] So cornerstone to that is Virtual Private Cloud, VPC. [57:31] That really allows enterprises to build these really nice [57:34] integrated hybrid environments. [57:36] It gives them a completely private virtual network running [57:39] inside of GCP. [57:41] It used to be that everything running on Cloud had public [57:42] IP addresses. [57:44] Now with VPCs you can have a completely private environment, [57:47] private IP addresses, private DNS, and full control. [57:51] And you want to be able to also control what applications [57:55] outside of your VPC can actually have ingressed in as well as to [58:01] make sure you have full control of anything running inside your [58:03] VPC that you allow to talk to the outside world. [58:06] And on top of all that, you always want full auditability [58:10] and full telemetry so that you really see everything that's [58:13] going on, even inside of these managed services. [58:15] They should not be opaque. [58:19] Also, you don't want application in data silos forced on you just [58:22] because of this move to public Cloud. [58:24] That was really one of several drivers behind the acquisition [58:28] of Apigee. [58:29] So with Apigee, you can actually put really elegant APIs in front [58:34] of technology so you turn them into consumable services and [58:39] that allows you to integrate applications in these services [58:43] whether they're built on Cloud and integrate them into [58:46] on-premise into your enterprise or on-premise and integrate them [58:49] into your Cloud applications. [58:52] So soon Chet Kapoor, the VP of Apigee, is going to be on stage [58:55] and he's going to go into far more detail on how they're [58:59] helping in building connected business platforms. [59:03] So Docker has been this amazing thing. [59:05] It's been this amazing gift to the industry what Docker has [59:09] created because it's really the first time that you've actually [59:13] been able to build these consistent application stacks [59:16] that run across hypervisors, different operating systems, and [59:19] different Cloud. [59:20] It's not perfect in terms of compatibility but it's really [59:24] setting us free to do some amazing things. [59:26] And it's because of Docker that Kubernetes is so successful. [59:31] Because now with Kubernetes, you need a control plane, you need [59:35] to be able to manage and orchestrate these [59:38] container-based environments and that's what Kubernetes does. [59:41] So it's really quickly become that de facto operational model. [59:47] And because it's open source, we're seeing customers, [59:50] enterprises run Kubernetes on-premise and then they're [59:53] running our managed service for Kubernetes, the container [59:56] engine, on Cloud. [59:57] It gives them this single operational model that's [60:00] entirely consistent across hybrid environments or they can [60:03] integrate them all together and run a single control plane on [60:06] GCP to even manage the on-premise world. [60:11] Serverless is a really important concept. [60:14] Developers shouldn't need to think about [60:16] managing infrastructure. [60:18] Servers should be provisioned automatically and just sized to [60:21] the workload. [60:22] And engineered right, it's more reliable, it's easier, it's [60:24] more efficient. [60:25] The spoiler alert is the servers are still there. [60:28] When you need them, they become plentiful. [60:30] And when you don't need them, they just go away. [60:33] It's really been a design approach. [60:35] Serverless is not new to Google. [60:36] It's been a design approach across many of our [60:38] major services. [60:39] On the compute side, App Engine and Container Engine [60:42] are serverless. [60:43] On the database side, Datastore and Spanner you don't see [60:45] infrastructure, and also with BigQuery. [60:49] Each service consumes no servers when there is no load and they [60:52] all scale out horizontally for when you need more horsepower. [60:58] And today we're announcing the beta of our newest serverless [61:01] offering - Cloud Functions. [61:08] So functions are simply fragments of code and they [61:11] get applause. [61:12] That's great. [61:13] But what developers do with these fragments of code is they [61:15] connect services together and they plug them into a growing [61:19] corpus of events across GCP and that's how they tether it in. [61:24] But what it does is powerful. [61:26] It lets you take this generic Cloud that's good for everybody [61:30] and you can personalize it in a way that is meaningful to you. [61:34] During the alpha, we saw people do some pretty amazing things. [61:37] One example that I love was one company actually looked at event [61:42] logs and then they keyed off certain events in the event logs [61:45] and they actually automatically filed bugs in JIRA. [61:47] I think that was pretty cool. [61:48] And then we also see people doing PII scanning. [61:51] They plug in for the PII they care about, and any time an [61:54] object comes into GCS, they can actually scan for it. [61:57] So this ability to be able to extend GCP in a way both for [62:02] developers as well as system administrators are [62:04] virtually endless. [62:06] So I mentioned earlier that back in 2008 Google App Engine was [62:12] really this pioneering serverless and it was ahead of [62:15] its time. [62:16] The core promise to developers still remains the same - bring [62:20] your code and Google will handle everything else. [62:25] It empowers applications to scale from one request per day [62:29] up to millions of requests per second. [62:31] And so when you actually liberate developers from [62:34] managing and patching servers, dealing with scale, dealing with [62:37] load balancers, version rollouts, managing databases, [62:41] they can create great things. [62:43] That's what's been happening at Snap, Home Depot, [62:46] and Philips Lighting. [62:47] And internally to Google, App Engine has been around a [62:49] long time. [62:51] So our corporate IT, we have thousands of App Engine-based [62:53] apps in production that Googlers depend on each and every day. [62:59] So beginning today, we're delivering a promise to Google [63:03] App Engine but we're making it available to entirely expanded [63:07] developer communities. [63:08] It's a focus on more openness, developer choice, and [63:12] expanded portability. [63:13] So out of the box, we now support seven popular languages [63:17] - Java, Ruby, Go, PHP, Python, C#, and Node - or you can now, [63:23] for the first time, bring your own runtime, bring your own [63:26] framework to App Engine Flexible Environment. [63:28] As long as it runs in the Docker Container, it now runs [63:31] on App Engine. [63:36] Thank you. [63:40] So I can't think of where serverless would matter more but [63:42] to mobile developers. [63:44] Do you know a mobile developer that wants to manage [63:46] operating systems? [63:47] What they want to focus on is building a great user experience [63:51] for their users. [63:52] So we've been working really closely with the Firebase team, [63:56] which is Google's mobile application development [63:58] platform, which supports iOS, Android, and Web. [64:02] They actually jut surpassed a milestone last month. [64:05] In the last 11 months, they now have one million active projects [64:10] on Firebase. [64:12] So, really amazing momentum. [64:14] Today we're actually making Firebase even more powerful [64:17] because we're integrating it closer so that you can access [64:20] GCP resources right from Firebase. [64:24] The first thing we did is we integrated Cloud storage into [64:27] the Firebase SDKs, so now you can access any GCS bucket [64:32] anywhere in the world right from Firebase. [64:34] So you get this direct to mobile upload and download for every [64:39] Cloud storage user from a mobile platform. [64:43] One thing that people love about Firebase is it has this amazing [64:46] built-in analytics capability and that's what developers use [64:50] to understand their users. [64:53] Now we've actually integrated that so you can actually take [64:56] the analytics data in Firebase and then drive it into BigQuery [65:00] and that allows even more analysis and really [65:03] understanding your users in realtime. [65:07] And we've also integrated the Cloud Functions that we just [65:10] talked about with Firebase, and so that allows Firebase [65:12] developers to extend their backend logic in crazy ways. [65:17] They can send push notifications, transform data to [65:19] machine learning, define custom business logic, all with just a [65:23] few lines of code. [65:24] In fact, the integration with Cloud Functions has been the [65:27] number one feature request of Firebase developers over the [65:30] last year. [65:33] And then finally, we love lawyers almost as much as we [65:36] love developers, so we're actually working to extend GCP's [65:40] terms of service to cover many of the Firebase products. [65:43] So we're not there yet but soon you'll be able to bundle much of [65:46] Firebase into essentially the same contract that you use [65:49] for GCP. [65:52] So we'll be making a lot more integrations over the [65:54] coming months. [65:55] Greg is coming back now and he's going to show you how we use GCP [65:59] and Firebase together to really modernize corporate [66:02] IT apps. Greg. [66:10] >> GREG DEMICHILLIE: Thanks, Brian. [66:12] Brian told you we're going to take an example of an existing [66:15] on-premise application and we're going to modernize it with [66:18] Google Cloud. [66:19] Now in this case, the app that Chris and I are going to work [66:22] with is an old ASP.net application. [66:29] Well that's not an ASP.net application. [66:37] That's an ASP.net application. [66:39] This is an application used by an insurance company. [66:42] So the adjusters go out in the field, they take their digital [66:44] camera, they take pictures of the accidents, they put it in [66:47] their SD cards, and then they upload the images into [66:50] the application. [66:52] It's also got an API on it so that you can manage all the [66:55] images that are associated with a given claim. [66:58] Now we're going to start bringing this application out of [67:00] the '90s and into the modern era. [67:03] We're going to start by giving our agents a good mobile [67:05] application so they don't have to lug a laptop around. [67:09] Firebase makes that super easy. [67:11] You can build a mobile application without having to be [67:14] a backend expert. [67:16] In this case, what Chris is showing you is the line of code [67:19] that integrates Firebase with Cloud Storage. [67:23] So this line of code allows the camera to take a picture, that [67:26] picture then automatically gets uploaded into a Google Cloud [67:30] Storage bucket. [67:31] So to show how this is going to happen, Chris is going to take a [67:33] quick photo with the phone. Yep. [67:39] >> CHRIS: All right, everyone, say cheese. [67:43] I love having audiences do that. [67:46] >> GREG DEMICHILLIE: So that photo now is being uploaded and [67:48] it's being stored into a Google Cloud Storage bucket. [67:51] Now to show that it really is in a GCS bucket, we're going to [67:53] switch over and we're going to list the contents of [67:55] this bucket. [67:56] Now the Windows users among you are going to notice that he's [67:59] using PowerShell. [68:01] That's because he's using Cloud tools for PowerShell. [68:03] It gives Windows developers a first class experience using [68:06] Google Cloud platform with the services that they love. [68:09] And sure enough, there is the picture that was just [68:11] uploaded today. [68:12] So now we've got a claim picture in GCS, how do we get it to this [68:17] old legacy application? [68:19] Apigee and Cloud Functions make that simple. [68:22] The good news is our legacy app has an API, as I mentioned. [68:26] The bad news is the app was written over a decade ago so [68:29] it's a SOAP API. [68:30] So it's not really very friendly to a modern restful developer [68:34] like Chris. [68:36] Apigee, however, is an Enterprise API [68:38] Management platform. [68:39] We're going to start by using it to create a proxy to convert [68:44] this legacy SOAP XML into modern JSON. [68:47] So here is the Apigee console. [68:49] We've connected it to the WSDL of the service. [68:52] I can't believe I'm saying that in 2017. [68:54] And on the left side, you see the SOAP XML, and on the right [68:59] side, Apigee has automatically created a much friendlier, [69:02] easier for the developer to use, JSON version of it. [69:06] Now Apigee has a lot of other capabilities. [69:08] So, in addition to doing that, Chris has applied a quota so [69:12] that we don't flood this poor service with millions of mobile [69:15] agents all trying to upload photos at the same time. [69:17] It automatically provides throttling. [69:20] We've also put authentication in so that only our agents and only [69:24] those mobile applications can upload the data. [69:27] Now Apigee has got tons more capabilities; I am only [69:31] scratching the surface. [69:32] But for our purposes, that's good enough for [69:34] this application. [69:35] So we've got images in a GCS bucket, we got our application [69:38] that's now got a nice modern API on it, how do we connect [69:42] the two? [69:43] Well I could deploy a VM, but that's overkill, right? [69:47] All I want to do is copy a little file from GCS to an API. [69:50] Why should I pick an operating system and have to patch the OS [69:53] and deploy a big, heavy VM image? [69:55] Cloud Functions allows me to just write a snippet [69:58] of Javascript. [69:59] This Javascript is listening to that GCS bucket. [70:02] So any time a file gets added, this snippet gets invoked. [70:06] And if he goes down, you'll see he is using the post method, [70:09] that nice, restful interface onto our web service. [70:13] So now when Chris takes a picture, we should see the [70:15] picture go all the way through. [70:16] Now I don't have a car crash on stage, Chris. [70:19] I think you've got though some sort of a next best thing [70:21] for us? [70:22] >> CHRIS: Actually, what I am prepared to do, Greg, is I've [70:26] got some scale model vehicle replicas here that I am going to [70:31] use to simulate a high speed collision scenario. [70:34] >> GREG DEMICHILLIE: This is scaled down in a digital [70:35] scale up. [70:36] Yeah, very good. [70:37] >> CHRIS: You ready? [70:38] >> GREG DEMICHILLIE: Yeah, go for it. [70:38] >> CHRIS: Vroom, crash. Ah! [70:47] >> GREG DEMICHILLIE: Please tell me you didn't misplace [70:48] the phone. [70:50] So he's going to take the picture of, oh, the humanity, [70:56] and it gets uploaded, and now he's going to switch back to our [70:59] application, our corporate app. [71:01] He's going to hit F5 because it's the '90s and we have [71:03] to refresh. [71:05] And there we go, there is our application. [71:10] We started with a legacy on-prem app, we used Firebase to build a [71:13] mobile application that didn't even require a backend expert, [71:16] we wrapped the old SOAP API in modern JSON, and Cloud Functions [71:21] was this wonderful glue layer to connect everything together. [71:24] But one more thing. [71:25] This application started as an on-prem application and we've [71:28] wrapped it with all this amazing Cloud stuff, but it's still [71:31] sitting on-prem. [71:33] Let's fix that bug, too. [71:34] With just a couple clicks, Chris will start the Migration Wizard. [71:40] In about 45 minutes, we will live migrate this running [71:43] service into Google Cloud platform where it runs on [71:45] Google's amazing infrastructure, [71:47] network, operations, and reliability. [71:49] And now we've really taken this app and brought it into the [71:51] modern era. [71:52] Now I'll turn it back over to Brian. Thanks. [71:59] >> BRIAN STEVENS: Thanks, Greg. [72:02] So, great companies use data to react faster to the market, [72:07] build novel new products, it even changed culturally how [72:10] they work. [72:11] Breaking down data silos to see bigger and better pictures of [72:14] what's going on in realtime is really super important. [72:18] And so for GCP to enable that, we deliver an end-to-end managed [72:21] analytics platform. [72:22] It now spins storage, data warehousing, ingestion, you saw [72:26] PII cleaning, ETL, batch and streaming modes, visualization, [72:31] and even machine learning. [72:33] It is expanded and extended by a great set of partners. [72:38] So Google is actually pretty good, right, about being a [72:41] data-driven company. [72:43] But to support that, it was all about Dremel. [72:45] Dremel was the purpose-built analytics engine inside of [72:49] Google that we depend on. [72:51] There is a white paper out on it. [72:53] It surfaced as Google Cloud BigQuery for users outside [72:57] of Google. [72:58] It's great because it scales horizontally in realtime. [73:01] It can ingest millions of rows. [73:03] It can process trillions of rows a second. [73:06] And you really need that for a lot of new applications to have [73:10] realtime decision-making. [73:11] Think about like what's happening with data ingestion [73:13] and IoT and social. [73:15] So it's great for realtime results across ever-changing [73:19] data sets. [73:21] As part of that, it's a big part of our data analytics platform. [73:25] What we've been doing is we've been actually connecting this [73:27] sophisticated data analysis that's inside of BigQuery with [73:32] rich data sources such as advertising platforms. [73:35] So today, to make that easier, we're now seeing the Google [73:39] BigQuery Data Transfer Service. [73:41] What that does is it automates the transfer of data from SAS [73:46] applications into BigQuery but it does it on a scheduled, [73:49] managed basis. [73:51] So today we have connectors for AdWords, [73:54] DoubleClick, YouTube Analytics. [73:57] And then once the data is inside of BigQuery, that's where [73:59] further enrichment happens. [74:00] You can integrate the ads data with weather data, geo data, [74:04] your sales data. [74:06] So it makes it really easy for marketing teams to become [74:09] empowered and they can build marketing analytics warehouses [74:12] on GCP. [74:15] But the process of actually bringing data into a data [74:18] warehouse can still be cumbersome. [74:19] I've seen stats that these data scientists spend 75% of their [74:22] time just like dealing with these disparate data sources and [74:25] cleaning them up so that they can be homogenous enough to [74:28] relate to each other. [74:30] And we want to make that easier. [74:32] That's why today we're introducing Cloud Dataprep. [74:34] Dataprep is this intelligent data service that allows you to [74:38] visually explore and clean your data so that it can be [74:41] integrated into a BigQuery environment. [74:44] It's as simple as using a mouse cursor. [74:46] You can hover over an attribute in a JSON object and you can [74:50] decide that you want that attribute to be a top level [74:52] column in BigQuery. [74:54] Or we've seen people take a single field that has a location [74:57] address and the Dataprep service allows them to break that apart [75:01] and say I want separate fields, separate columns for state and [75:05] Zip Code. [75:06] And it's really smart. [75:07] It actually uses machine learning itself, because often [75:09] in data, there is data quality issues as well. [75:12] And so based on machine learning, it suggests [75:14] transformations to your data to make it cleaner. [75:20] So all of these need not be one time ingestions and [75:23] transformations orchestrated under a Cloud data flow. [75:26] The aim is that you actually get out a batch and you're building [75:29] a streaming data analytics pipeline. [75:33] We've integrated with Data Studio out of Google's [75:35] analytics team. [75:37] What that now allows is you can actually visualize, create these [75:40] really rich dashboards and charts almost automatically. [75:46] Or you can integrate all your data that's now in Cloud with a [75:49] lot of our machine learning. [75:51] So you can actually train whole new models and then use them for [75:53] recommendations and predictions. [75:56] Oh, with all of this, never will you be managing infrastructure. [75:59] You won't be installing software. [76:01] You won't be integrating software. [76:03] You won't be managing performance and scalability. [76:06] You'll just get right to the mission at hand - [76:08] analyzing data. [76:09] So, enough talk. [76:11] Greg is actually back to live demo again how all this works. [76:18] >> GREG DEMICHILLIE: Thanks, Brian. [76:20] This is my last of the four promised live demos. [76:23] Before we start, I would be remiss if I didn't thank Chris, [76:26] Neil, Martin, and Robert who have been helping with building [76:32] these demos. [76:33] Would you thank them for me before we get too far into this? [76:37] Thank you. [76:39] So to give you a picture of how this data platform puts [76:41] together, we're going to use a scenario of an advertising [76:44] agency that puts ads in the back of those screens in New York [76:47] City taxis, if ever you've been in a cab recently in New York. [76:50] There's an ad that plays and you can interact with it. [76:53] To give you the picture first, this is a simplified [76:55] architecture of this data solution. [76:58] The taxis are sending their position in realtime, where they [77:01] are in the city, as well as their destination, how many [77:04] passengers they have, and what sort of interactions the [77:06] customers are doing. [77:08] Pub/Sub is then ingesting that data. [77:10] We're using dataflow to process a data pipeline to handle and [77:13] process all that data. [77:15] We're also archiving the entire historical data in BigQuery [77:18] which is a really powerful tool for you to do ad hoc analysis of [77:22] all of your historical data. [77:24] And at the end, we want to make a visualization. [77:28] So why don't we actually see the pipeline in action? [77:33] So this is the dataflow visualization tool. [77:35] You see Robert has hovered. [77:36] We are ingesting a little over 15,000 new elements per second [77:41] into this database of all of this information. [77:44] But how do you visualize all that? [77:46] Google Maps is super powerful but I would hate to point 15,000 [77:49] updates per second at my Chrome browser. [77:52] So dataflow actually allows us to down sample that event stream [77:56] to give us a realtime visualization of taxi positions. [77:59] This is that visualization. [78:01] This is showing where all these taxis are in New York City at [78:03] any given time. [78:05] Now what if I want to look at one particular customer who has [78:08] placed ads with us, a particular agency? [78:10] We can filter that dataflow and now the visualization will [78:14] update so that we're only seeing ads that are being placed for a [78:18] specific store. [78:19] In this case, the Acme store, it looks like, is on the Upper [78:22] East Side. [78:23] And sure enough, we see that we're placing ads roughly [78:25] geographically near the store. [78:28] But right now we're just using a dumb rule for this. [78:30] It's just a very simple rule. [78:32] Where is the store? [78:32] Where is the taxi? [78:33] Place the ads. [78:34] Can we do better? [78:36] Can we automate this to start to take into account all the [78:39] various data that we have about this? [78:42] Let's start by using BigQuery to sort of explore this dataset. [78:46] The query Robert has here looks at all of the taxis coming from [78:51] the airports and tries to see how many of them are traveling [78:54] near restaurants. [78:56] Now note here he's querying this historical data even as we're [78:59] ingesting new data at 15,000 events per second. [79:02] So BigQuery gives him the ability to have an always [79:05] up-to-date realtime view into his data. [79:08] In this case, we see that roughly 17% of our taxis are [79:12] from the airport and near a restaurant. [79:15] That sounds like an insight that maybe we ought to do some ad [79:17] targeting there. [79:19] And if you look at this query, you can see it really is a [79:21] non-trivial query and BigQuery is handling all that for us. [79:25] So right now we are manually configuring these ads. [79:31] I think what we want to do is look at a way that we can do [79:34] this in an automatic fashion. [79:37] Now we could similarly search for every different customer, [79:40] but how do we build a generalizable and [79:42] flexible model? [79:43] That sounds like a job for machine learning. [79:45] Machine learning is tailor-made for taking large chunks of data [79:48] and building models that give you insights out of that. [79:51] So we're going to build a model that takes into account all of [79:53] the data we have. [79:54] Where did the taxi start? [79:55] Where did it end? [79:56] How many people were in it? [79:57] What time of day is it? [79:58] Where is it going? [80:00] First, we're going to start by using, as you see in this [80:03] picture, we're going to use Dataprep to take all that [80:05] historical data, clean it up, make sure that it's ready to [80:08] use, and then we're going to use it with ML Engine. [80:11] So this is Dataprep. [80:12] The first thing to notice is, across the top, Dataprep has [80:15] automatically inferred the schema from my table. [80:18] I didn't have to tell it that. [80:20] Now if you look at that passenger count column, you'll [80:23] notice that it looks a little funny. [80:25] The histogram up at the top includes some elements where it [80:29] says there are 10 or 11 people. [80:31] Now I don't know about you, but I've been in a New York City cab [80:34] and I'm pretty sure you can't actually put 10 or 11 people in [80:36] a cab. [80:38] So that's a common data bug. [80:40] Dataprep allows me to fix that in a number of ways. [80:43] I could delete the data. [80:44] I could replace the 11 with a 1 on the theory that it's a fat [80:48] finger mistake. [80:49] And Dataprep allows me to build simple recipes that clean up my [80:52] entire data at scale. [80:55] It also can do things like if you look at our location column [80:57] there, it's a composite column. [80:59] It's longitude and latitude comma separated. [81:02] I would really like to have a latitude column and a [81:04] longitude column. [81:06] Dataprep allows me to build a recipe that splits columns with [81:09] composite data into separate columns so that I now have a [81:12] better schema to work with. [81:14] Now this just scratches the surface of how Dataprep can take [81:17] this really tedious job of preparing data for analytics and [81:21] make it much faster and must more reliable. [81:23] Once we've done this, machine learning, as you heard from Faye [81:26] Faye, allows me to build a model on my laptop, upload it to [81:30] Google, and deploy it at scale. [81:32] Let's see how that actually works. [81:34] If you scroll down in the dataflow pipeline, you'll see [81:36] we've added an ML engine call. [81:39] So, now as our taxi data is coming in, we're calling to [81:42] machine learning with the data and saying help us choose the [81:45] best ad to target for this particular user. [81:48] So let's switch back over to the map with the deployed machine [81:51] learning model and now you'll see that we start seeing some [81:55] interesting combinations. [81:57] When there is one passenger coming in towards Midtown, the [82:00] taxi is currently in Chelsea and it's on its way to Midtown, [82:04] we're presenting an offer for Manhattan Bagels because it's a [82:06] single person, maybe he or she wants a bite to eat. [82:09] If we pick a taxi that has more than one person in it - if you [82:20] have a taxi with more than one person in it - there it is - in [82:23] this case - that's still not right. [82:25] That's a one person one. [82:26] But they're heading into a different location and instead [82:29] we've chosen a winery. [82:30] If you pick a taxi with three people, the system would [82:33] automatically recommend discounted tickets to Broadway, [82:35] for example, as multiple people coming to Midtown in [82:38] the evening. [82:39] So that's Cloud ML making suggestions based on all our [82:43] available data. [82:45] We used Pub/Sub to ingest a huge stream of data at scale. [82:49] We used dataflow to build a pipeline to give us a good [82:51] analytics system. [82:53] We used BigQuery so that we always had the complete [82:56] historical data available. [82:57] We never had to deal with sample data. [83:00] And at the end, we used Cloud ML and Dataprep to build and train [83:03] at scale and model. [83:05] What's important here is what you didn't see. [83:07] I didn't deploy a virtual machine. [83:09] I didn't deploy patches. [83:11] I spent my time actually in the data. [83:13] That's the point of Big Data is to spend the time in the data [83:16] and not in the machines that take care of the data. [83:19] That's what Big Data on the Google platform does. [83:22] So I want to thank you all and I'll turn it back over to Brian. [83:29] >> BRIAN STEVENS: Thanks, Greg. [83:31] So people that know me know that I loathe buzzwords, and so I'm [83:35] probably going to offend a few people in the audience, but [83:37] digital transformation is absolutely one of the ones that [83:40] I hate. [83:41] I think in part it is because I've never seen even two people [83:43] even agree on what it even means. [83:45] But what I do know is this, is that markets are incredibly [83:48] competitive, each year more so than past, and the new companies [83:52] that actually take advantage of new technology without having [83:55] any legacy move really quickly to either disrupt or create [83:59] whole new lines of business. [84:01] The companies that win are going to be the ones that shed the [84:04] mundane, take advantage of state-of-the-art technology, and [84:08] actually put people to work, their intellectual horsepower, [84:12] and they're creating new end-user value for [84:13] their customers. [84:15] That's what we want to do at GCP is enable you on that journey. [84:18] So thank you and it's my pleasure to introduce Prabhakar [84:21] Raghavan, the leader of our G Suite team. [84:31] >> PRABHAKAR RAGHAVAN: Thanks, Brian. [84:32] It's so exciting to be here. [84:38] What is productive work? [84:41] It's a systematic progress from a huge multitude of choices down [84:46] to a decision or an outcome. [84:49] Whether it's figuring out where to file the document that I just [84:53] got, choosing the words for my next email, or at the epic end, [85:01] choosing all the pieces that come together in a grand [85:03] symphonic work, in every instance we are going from chaos [85:08] and ambiguity down to an outcome. [85:13] You will notice that the three examples I just gave you range [85:18] from the mundane to the sublime. [85:21] The mundane end was where do I file the document. [85:23] The sublime end was the grand symphonic work. [85:29] Machines are really, really good at taking care of the [85:31] humdrum down there, but it takes humans to do the truly [85:36] creative work. [85:39] At G Suite, we are obsessed with the idea that computers should [85:44] constantly raise the bar on what they can get done at the mundane [85:50] so that our users, your employees, have more and more [85:54] time to focus on truly creative work. [86:00] In a study of several million of our Gmail users today, we find [86:04] that one in eight of their email replies is actually [86:08] machine-generated and they take those machine-generated replies [86:12] and send them off and they're good. [86:16] So that's a case where we are raising the bar on what [86:18] computers couldn't do five years ago but today they can take care [86:23] of that and leave humans to focus on truly creative work. [86:28] And so what that means is while our computers may not have [86:33] written Beethoven's Ninth Symphony, maybe we could have [86:36] freed him up to write nine more symphonies. [86:44] At G Suite, this has been an obsessive pursuit for over [86:47] ten years. [86:48] Today, three of our apps are on over a billion smart phones. [86:55] We have over three million paying businesses that use the [87:00] G Suite. [87:02] But I will say this has been an evolution, a journey for us, [87:08] because we have gone from applications crafted for [87:12] consumers and then outfitted for enterprises after the fact to [87:19] having an enterprise first focus. [87:22] I'm going to give you a couple of data points for that. [87:26] One thing, we have begun to do early adopter programs with our [87:30] best enterprise customers so that we don't just build a [87:35] product and push it out the door. [87:37] We work closely in the final months of development with our [87:40] top customers and partners. [87:43] We take their feedback and refine the product and get it [87:47] just right before we deliver it. [87:50] Here are three examples of recent early adopter programs [87:57] that we have done. [87:58] I call these out for a specific reason. [88:03] Each of these apps is something that was built for [88:08] enterprises only. [88:09] These were not consumer apps ported to enterprises. [88:13] There is no consumer pedigree here. [88:15] The Jamboard, App Maker, Google Cloud Search, every one of these [88:19] is an enterprise-only product. [88:20] You will see the Jamboard in a bit. [88:23] We don't expect too many consumers to be buying those. [88:28] Now at the center of productive, collaborative work is content. [88:36] I am thrilled to share a statistic with you today. [88:39] Google Drive today has over 800 million monthly active users and [88:46] it's on a tear to hitting a billion. [88:49] It will soon become Google's latest billion user product. [88:56] We think of Google Drive as the premiere personal storage [89:01] solution, but we also see an opportunity there where it can [89:08] evolve from being personal storage to serving the needs of [89:12] the enterprise for file sharing. [89:17] And as we go through this journey, I will be making five [89:20] announcements now to represent steps along the way. [89:27] First, Team Drives goes into general availability today. [89:40] As you can tell, this has been one of the most demanded [89:42] features from our customers. [89:45] It lets teams easily share content in their drive and [89:50] manage the sharing. [89:53] Now once the content is in a drive or in a team drive, we [90:00] would like to make sure that the enterprise needs of compliance, [90:05] of archiving, litigation holds, e-discovery, all of that stuff [90:08] is taken care of. [90:09] So I am pleased to announce general availability for Google [90:13] Vault for Drive content. [90:15] So once the content is in the drive - okay, I see a few people [90:19] that are excited about that as well. [90:21] Thank you. [90:23] Google Vault for Drive has arrived, general availability. [90:28] All right, you say. [90:29] But if you're a prospect, you look at it and go that's all [90:33] very well for content in the Cloud but most of my content [90:38] sits in on-premise file servers. [90:40] What do I do about that? [90:42] I'm excited to announce that we are acquiring [90:45] Vancouver-based AppBridge. [90:48] AppBridge is a company that builds connectors from [90:52] on-premise file servers to siphon the content up into [90:55] Google Drive. [91:03] The fourth announcement I am making today around Google Drive [91:08] has to do with the following. [91:10] Once the content is up there, how do you make it accessible [91:13] and easily manageable for someone who has got a Windows [91:17] laptop or a Mac? [91:20] No more sync clients. [91:21] No more checkboxing which files you want to save. [91:23] No more worrying about how much hard drive space do you have. [91:28] Drive File Stream takes care of all of that seamlessly and [91:31] obliviously so you don't have to worry about all of these [91:33] minutiae and can work as if Google Drive, the entire Cloud, [91:38] is connecting to your laptop. Great. [91:43] Now once all of that content is up in the Cloud, it opens up the [91:48] potential for Google's machine learning magic. [91:55] Think of the following. [91:56] You come to your favorite file depository and you know there is [92:00] a file you're looking for but you're scratching your head did [92:03] I create it, did Joe share it with me, what key words does [92:07] it have? [92:08] In our studies, you spend something like 40 seconds on [92:11] average trying to figure out the right filters and restrictions [92:15] and browsing and keywords before you actually get to the file. [92:20] You shouldn't have to. [92:21] There is one file you are looking for. [92:23] Google's machine learning magic will build a predictive model [92:28] for who you are and your activity and serve up that file [92:30] before you even ask for it. [92:32] We call this Quick Access and today it is generally available [92:35] for both Drive and Team Drives, Android, iOS, and the Web. [92:42] So that is the cache of announcements I had to make [92:46] around Google Drive. [92:53] Now that's all about content, but really where do the people [92:58] come in here? [93:02] So for this next segment of announcements, rather than my [93:05] keep talking, we're going to do a bunch of demos around teams [93:09] and meetings for you. [93:11] I'm going to call on stage my colleagues Scott Johnston and [93:15] Jonathon Rochelle. [93:26] >> SCOTT JOHNSTON: I'm Scott Johnston and I'm going to show [93:28] you how we rebuilt Google Hangouts with a focus on making [93:32] teams productive. [93:34] And to do this, we're going to visit a company called Cloudy [93:37] Coffee which is in the midst of launching a coffee bean with [93:40] 100X the caffeine of a normal bean. [93:43] This is a bean that I desperately needed this morning [93:46] and I think will be very popular in the market. [93:49] We're going to start by looking at the new Hangouts chat, [93:52] completely rebuilt to be an intelligent messaging app [93:55] for teams. [93:57] Now, Cloudy and our G Suite customers already use Hangouts' [94:01] direct messages every day to keep work moving forward. [94:05] But in the new Hangouts chat, we've added rooms. [94:09] Rooms are a central place for team and project discussions. [94:14] To help look at this and help us walk through this, I've got [94:16] Mandy at the controls. [94:18] >> MANDY: Fired up and ready to go. [94:20] So I'm going to bring up Cloudy Coffee Room in the new [94:24] Hangouts chat. [94:26] >> SCOTT JOHNSTON: So the first thing you will notice in this [94:28] room is that discussion, realtime discussion is threaded, [94:31] and this allows me to separate the lunch conversation from the [94:35] work and allows teams to dive deeply into discussion without [94:40] fragmenting other discussions. [94:43] Mandy, why don't we move a bit lower in this room and look at [94:46] some other stuff? [94:47] So here we see Nicole has posted a brochure for our launch. [94:54] That brochure is in Docs. [94:57] In the new Hangouts chat, Docs and Drive are deeply integrated [95:01] and the room manages the permissions for you, so anybody [95:03] who is a member of the room now or in the future always has [95:07] access to work with that file. [95:10] So now we have a central place to discuss project and [95:12] team information. [95:14] We know that search is critical, and so we have built a powerful [95:18] search directly into the Hangouts interface. [95:22] >> MANDY: So on top of free text search, you can also filter by [95:25] people or types. [95:26] So you can always find the content in your rooms. [95:28] Let me take a look at slides. [95:30] Ah, and there's the sales forecast that Patrick put in. [95:34] >> SCOTT JOHNSTON: Perfect. [95:36] Teams work with a myriad of tools today, and so we have [95:40] created the Hangouts platform that lets third parties deeply [95:44] integrate with Hangouts and Team rooms. [95:48] The platform supports a wide range of capabilities from [95:51] lightweight scripting with our Google Apps script so you can [95:55] automate team workflows quickly all the way through to [95:59] intelligent bots. [96:01] So let's look at an example. [96:03] Cloudy Coffee uses a sauna, a work tracking product, to stay [96:09] on top of their launches and know who is working on what [96:12] and when. [96:15] >> MANDY: Okay. [96:16] So right from inside the room, I can assign a task to Mike, who I [96:28] like to assign all my tasks to. [96:30] And with one simple click, task is created. [96:34] >> SCOTT JOHNSTON: Great. [96:36] We've teamed up with a number of companies like Zendesk, [96:38] ProsperWorks, Box, and more, to deeply integrate their products [96:43] into the Hangouts platform. [96:45] And let me show you actually how we're using the platform itself [96:48] to integrate our own products. [96:51] We've built an intelligent bot we call Meet. [96:54] Meet uses natural language processing and machine learning [96:58] to automatically schedule meetings for your team. [97:01] >> MANDY: I'll ask Meet to find a time for us - for the people [97:03] in the room today. [97:04] >> SCOTT JOHNSTON: Perfect. [97:05] >> MANDY: Oops, maybe I'm - I may have had too much of [97:08] that coffee. [97:09] >> SCOTT JOHNSTON: Too much of the coffee? Yeah. [97:16] >> MANDY: Oops. [97:24] >> SCOTT JOHNSTON: So what's happening here is Meet is going [97:26] to go out, it's going to look at all our calendars for members in [97:28] the room, it's going to find the optimal time for us to meet this [97:31] afternoon, and automatically book it in Google Calendar. [97:35] >> MANDY: Actually, let's move that meeting to tomorrow. [97:38] >> SCOTT JOHNSTON: Okay. [97:40] So here we are seeing with simple conversational commands [97:42] we can do something that otherwise used to take many, [97:45] many steps. [97:47] Wait, if you're moving that meeting to tomorrow, that means [97:50] all of you are going to have to stay until tomorrow for the [97:53] demo. Is that okay? [97:54] Are you guys good with that? [97:55] I can lead a sing-along. [97:56] We have snacks. No? [97:58] >> MANDY: Good point, Scott. [97:59] Let me just move that to right now. [98:00] >> SCOTT JOHNSTON: Okay. [98:01] So we're going to ask Meet to schedule a team meeting [98:03] right now. [98:04] And before we jump into this meeting, let me talk about [98:07] meeting technology today. [98:09] I look around and I see us landing rockets on rafts in the [98:13] ocean and it's still so hard to get people into a meeting. [98:17] Our customers sometimes spend ten minutes getting a [98:19] meeting ready. [98:20] Somebody doesn't have an account, the meeting system sent [98:24] them 72 phone numbers and you can't find the code. [98:27] And so this is what we obsessed about when we rebuilt the [98:29] Hangouts meeting experience. [98:32] What does it mean to obsess about it? [98:34] It means no plug-ins required. [98:35] One click and you're in. [98:37] We dramatically reduced the code size and optimized the [98:42] experience so that you're instantly in the meeting, you [98:45] have less CPU fan, and your video and audio quality are [98:49] dramatically improved. [98:51] All right, I'm done with my rant. [98:53] >> MANDY: Great. [98:54] I think we have the entire team already one. [98:55] >> SCOTT JOHNSTON: All right, great. [98:57] >> MANDY: So when you click in, you're instantly taken to what [98:58] we call the green room where you can check out and make sure [99:01] you're ready for the meeting and then join. [99:06] Hey, guys. [99:07] Welcome to the keynote. [99:08] >> TEAM: Hey. [99:11] >> SCOTT JOHNSTON: Here we are. [99:12] So what you're seeing here is the new Hangouts Meet, our [99:15] enterprise solution for video meetings. [99:18] So, sure, it loads fast. [99:21] All right, yeah. [99:25] So it loads fast, it performs well, and I'm sure you'll get to [99:28] try that when we launch it today. [99:30] But there's also friction getting into meetings in [99:33] other ways. [99:35] What about that consultant that doesn't work for the company, [99:37] doesn't use G Suite? [99:39] I don't know why you wouldn't use G Suite, but some people [99:41] don't I hear. [99:42] No problem. [99:43] Vroon is a consultant that is helping us with this coffee [99:46] campaign, and with a single link he has jumped in. [99:50] >> MANDY: So I've just accepted Vroon's knock and he's [99:51] instantly there. [99:53] Hi, Vroon. [99:54] It also looks like Mike has joined from the road. [99:57] >> SCOTT JOHNSTON: And that's because Hangouts Meet with every [100:00] meeting now can contain a dial in code so that you can connect [100:03] and participate in the meeting even when you don't have a [100:06] data connection. [100:12] So there is a lot more I could talk about with Hangouts Meet, [100:15] but what I want to stress is that the focus was to really, [100:18] really get you into the meeting as quickly as possible so the [100:21] technology would get out of the way and you could focus on [100:23] real work. [100:24] I don't want to spend more time taking because we're in [100:26] a meeting. [100:28] So now that we're actually in the meeting, why don't we talk [100:30] about the ad campaign for this new bean? [100:33] >> FEMALE SPEAKER: Yeah, I have some ideas that I can white [100:35] board but I don't know if you guys are going to be able to [100:37] see it. [100:38] >> JONATHAN ROCHELLE: Stop. [100:39] Please stop. [100:41] You said white board? [100:42] >> SCOTT JOHNSTON: Yeah, is there a problem? [100:43] >> JONATHAN ROCHELLE: And you're trying to reduce friction in [100:44] the meeting? [100:45] That's a great idea. [100:45] >> SCOTT JOHNSTON: What's the issue? [100:46] >> JONATHAN ROCHELLE: No, really. [100:47] I mean, it's 2017. [100:48] Nobody's going to be able to see it. [100:49] And a five person meeting is bad enough. [100:52] We've got 5,000 people here; they're not going to be able to [100:55] see the white board. [100:56] Oh, wait. [100:57] Actually, we can just have you point the camera to it, right? [100:59] Yeah, we'll do that. [101:00] >> SCOTT JOHNSTON: Yeah. [101:01] >> JONATHAN ROCHELLE: And you can't save it. [101:02] There's no way you're going to be able to save it. [101:03] >> SCOTT JOHNSTON: Can't I write don't erase? [101:04] >> JONATHAN ROCHELLE: Yeah, how about that? [101:05] Write do not erase on it. [101:05] That will work. [101:06] >> SCOTT JOHNSTON: Let me guess. [101:07] >> JONATHAN ROCHELLE: Or snap a picture. [101:08] >> SCOTT JOHNSTON: This thing they rolled out here, do you [101:09] have a better solution? [101:10] >> JONATHAN ROCHELLE: I think we might. Yeah. [101:12] Let's talk about that. [101:14] So what we need is not to waste the time once we're in the [101:17] meeting, right? [101:17] We got there really easily. [101:18] Let's not waste that. [101:19] Let's get work done. [101:20] That's why we meet, to collaborate and to work. [101:23] But there's never been a great tool for that. [101:25] The white board is close. [101:26] What we need is all the virtues of the white board. [101:29] It's fast, it's simple, and it's freeform, but we want that with [101:34] something where all the friction is removed. [101:37] That's why we created Jamboard. [101:40] Jamboard is a white board in the Cloud, in the meeting room, and [101:44] beyond your meeting rooms. [101:46] You just pick up the stylus and you start thinking, [101:48] communicating, and working with your team. [101:51] So let's see what that really means. [101:53] So T.J. is going to write some [101:54] ideas that we've talked about for this [101:55] new Cloudy Coffee ad campaign, but T.J. [102:00] won't be working alone here. [102:02] >> FEMALE SPEAKER: Hey, guys. [102:03] Now we can't see the Jamboard. [102:05] >> JONATHAN ROCHELLE: All right. [102:06] I think we've got a solution for that, though. [102:07] So wait for it. [102:09] The Jamboard knows that there is a meeting going on and [102:11] automatically presents a prompt, and with one click T.J. [102:15] can join the meeting and present the Jamboard to the meeting. [102:19] So everybody remote, everybody on the meeting can see what he's [102:21] doing on the Jamboard. [102:25] Hey, Patrick. [102:30] Patrick, you guys can see the Jamboard now, right? [102:33] >> PATRICK: Yeah. [102:34] >> FEMALE SPEAKER: Yep, thanks. [102:35] >> MALE SPEAKER: We're all good. [102:36] >> JONATHAN ROCHELLE: And by the way, we have an opinion on the [102:38] stylus that T.J. is using. [102:39] The Jamboard stylus is passive. [102:42] You don't have to charge it. [102:43] You don't have to pair it. [102:44] You don't have to dock it. [102:46] And when you lose it, it won't cost an arm and a leg because [102:48] it's passive. [102:50] It's intuitive and simple. [102:52] Most importantly, the Jamboard still knows the difference [102:54] between that stylus and your finger so you can write with the [102:58] stylus and erase with your finger naturally [103:00] and intuitively. [103:07] So, guys, why don't you help us out here? [103:09] Help T.J. out and take a minute [103:11] and let's brainstorm some activity. [103:13] So you know the ad campaign that they're working on, sometimes [103:15] you use those sticky notes if you're in the same room. [103:18] So the team is going to use the Jamboard companion apps and [103:22] anyone on the team can now add content and help T.J. [103:25] with his brainstorming. [103:26] So whether it's from their lap or their phone or their tablet, [103:28] from an Android or an iOS device, the companion apps let [103:32] them participate and get work done to add their ideas, to make [103:36] a point, to organize what is already there. [103:38] It's the same magical collaboration experience [103:40] hopefully you've gotten used to in Google Docs. [103:44] Now we're all at the white board. [103:45] So Lucy in Liverpool or in the room over there, or Bereen in [103:49] Brazil, or Tae in Taiwan, this is teamwork. [103:52] This is actual work happening. [103:54] Progress here and now while we're in this meeting. [103:57] So let's keep working. [103:58] We're looking at where to target this ad campaign. [104:00] So let's see another feature. [104:03] We're going to look at a billboard location. [104:05] So how about New York City? [104:06] Let's go big. [104:07] I've seen a lot of billboards in San Francisco [104:09] this week, actually. [104:10] But we're going to go big and go in Times Square. [104:12] So the power of the Web, the power of Google is right there [104:15] for T.J. or anybody using the companion apps, finding [104:18] relevant, functional, beautiful information to add to the Jam [104:22] and your Jam comes to life. [104:24] So T.J. is using Search and [104:25] Maps on the Jamboard to add useful content [104:28] for the team, and Bereen and Tae and Lucy and others are adding [104:32] other content. [104:33] So they're working on visuals. [104:35] What's the branding? [104:36] What is the branding feel we want for 100X caffeine coffee? [104:40] I'm afraid to ask. [104:43] So they're all using Search. [104:44] And Jamboard is also integrated with Google Drive. [104:47] So let's try something else. [104:49] You see somebody actually added using the companion app a slides [104:52] deck that we're working on the financial model for [104:55] Cloudy Coffee. [104:56] So spreadsheets, presentations, and documents, and anything from [105:00] Drive can be added to the Jamboard from the companion apps [105:04] so the team can see it, work on it, [105:06] and keep it. Excellent. [105:10] Now imagine we had more than one Jamboard. [105:16] Imagine we had more than one. [105:17] And we actually do. [105:18] We have actually somebody in New York on this meeting, one of our [105:21] best graphic designers. [105:22] I think we'll work on a custom logo for this Cloudy Coffee. [105:26] So Elon is working on a Jamboard out of our New York office. [105:29] You can see actually the avatars - T.J., if you could open up [105:33] that frame organizer - you can see the avatars of where [105:35] everybody is while they're working on the different frames [105:39] in the Jam. [105:40] And Elon is working on something and we're seeing it change [105:42] live here. [105:44] Everyone that's on the companion apps is seeing the same thing. [105:50] Thanks, Elon. [105:50] Awesome work. [105:57] So now I think we're pretty much ready to send this off to the [106:00] ad agency. [106:01] So T.J.'s handwriting is not actually the best. [106:03] That's probably the best I've ever seen, T.J. [106:05] But let's make it a little better. [106:07] The power of the Web and Google and the machine learning [106:10] capabilities of Google are available for T.J. [106:13] to make his handwriting really good too. [106:27] Excellent. And when you're done, [106:28] please don't snap a picture of this. Okay? [106:32] You don't need to. [106:33] Everything from Jam is saved in Google Drive so you can pick up [106:35] the progress next time. [106:38] That was just a brief overview of the new Hangouts chat, [106:42] Hangouts Meet, and the new Jamboard, built with the [106:45] enterprise power of security, intelligence, and scale that you [106:48] would expect from G Suite and designed to bring teams together [106:51] in several ways. [106:52] If you want to learn more, please participate in our [106:55] breakout sessions or come see us in the sandbox. [106:57] Back to you, Prabhaker. [107:05] >> PRABHAKER RAGHAVAN: Awesome job, Jon and Scott. [107:07] Thank you for that. [107:11] The final announcement we are going to make today thinks of G [107:16] Suite kind of in the same way you think of your teams. [107:19] They work really well together but they also need to work [107:23] outside of the team. [107:27] We decided that it was time for Gmail to have add-on capability [107:33] because people were wasting simply too much time going from [107:38] their primary work surface, which could be Gmail, digressing [107:42] over to a distraction, and then by the time they come back 20 [107:45] minutes are gone. [107:48] Rather than my doing the talking and explaining this, I'm going [107:52] to bring on stage a partner, Intuit, who will explain what [107:58] they did with Gmail add-ons. [108:00] Please welcome to the stage EVP and CTO of Inuit, [108:04] Tayloe Stansbury. [108:15] Welcome, Tayloe. [108:17] Take it away. [108:18] >> TAYLOE STANSBURY: Thank you, Prabhaker. [108:19] And good morning, everyone. [108:20] I am delighted to be here. [108:22] If I could just say a word or two about Intuit, our primary [108:25] brands are TurboTax and QuickBooks. [108:28] And with QuickBooks, we have some 1.8 million online and [108:32] mobile users around the world. [108:35] Now it turns out that about a half million of those users also [108:39] use Gmail. [108:40] It seemed natural that with this add-on capability we would want [108:44] to integrate the two so that those users could have a more [108:47] smooth flow between those applications. [108:52] So let me give an example. [108:55] Here we have a customer whose name is Craig. [108:56] He's a gardener. [108:58] He really loves making people's gardens beautiful. [109:00] And he's sitting in a cafe and he's reading his Gmail on his [109:03] phone and up comes a message from his customer Sarah who says [109:06] she's delighted with the work that he's done on her garden and [109:10] she's wondering how she should pay him. [109:14] He's thinking, wow, this is an opportune time to invoice her. [109:17] Now what it used to be is that he would have to jump out of [109:20] Gmail, go into QuickBooks, log in, find the place to do [109:24] an invoice. [109:26] But instead what we did with this add-on integration is that [109:29] we have a QB icon that is at the bottom of his Gmail. [109:33] He simply clicks on that, it infers context, single signs him [109:38] on to QuickBooks, drops him right into the place where he [109:41] can fill out an invoice. [109:43] And as you can see, it is excerpted from the email the [109:47] name and address of his customer. [109:49] Now all he has to do is fill in the rest of the details for his [109:52] invoice and off he goes. [109:55] Remember, he's still in Gmail. [109:58] He's never left the application and he's been able to fill out [110:01] the invoice entirely in there. [110:03] Boom. He sends the invoice and he's done. [110:06] It was that easy. [110:09] Now when he's done with that, he can actually quickly look at how [110:13] his invoices are doing with his other customers in QuickBooks [110:16] still inside the Gmail app, see how that's going, and get right [110:20] back into Gmail so he can finish his coffee and go off to make [110:23] other people's lawns beautiful. [110:26] So, to summarize, it was really easy to do this integration. [110:31] We wrote once and deploy many across Android, iOS, [110:35] and the Web. [110:37] We can't wait to get this to market. [110:39] It will be later this year. [110:41] And this is just one of many integrations that we expect to [110:43] do between QuickBooks and the G Suite. [110:46] We already announced one earlier around the calendar and we [110:50] expect to have more over time. [110:52] Thank you. [110:53] >> PRABHAKER RAGHAVEN: Outstanding. [110:53] Outstanding, Tayloe. [110:54] Thank you for being such a great partner in this. [111:01] All right. [111:03] I'm just going to finish with a wrap-up of the announcements we [111:06] saw today. [111:08] So here are the announcements we went through today. [111:11] Team Drives, general availability; Vault for Drive, [111:15] GA; AppBridge, Vancouver-based company, acquired for building [111:21] connectors; Drive File Stream and early adopter program; and [111:28] Quick Access for Team Drives across all platforms. [111:33] We saw the new Hangouts chat that Scott demonstrated. [111:36] That goes into early adopter program today. [111:40] Hangouts Meet in general availability. [111:43] That's the video conferencing piece. [111:46] The Jamboard. [111:47] We will begin taking orders soon and we should be generally [111:51] available in May at a price point just below $5,000 for the [111:56] board with an annual subscription fee of $600 for [112:01] the service. [112:03] And finally, you saw Tayloe showcase the Gmail add-ons, and [112:07] we've been working with a bunch of other partners to build out [112:10] exciting new add-ons. [112:11] That goes into development preview and we hope many of you [112:14] will come and join us on that exciting journey. [112:22] With that, I want to thank you for all your attention. [112:26] Next, welcome my colleague Chet Kapoor. [112:37] >> CHET KAPOOR: Thanks, Prabhaker. [112:42] Hello, how is it going? [112:44] >> AUDIENCE: Good. [112:45] >> CHET KAPOOR: That was a little weak. [112:47] I just wanted to let you know that we have extended [112:49] the keynotes. [112:50] We are going to have another two hours, so you should just get [112:52] really comfortable. [112:54] This is going to take a while, so you might as well enjoy [112:58] the show. [113:01] My name is Chet Kapoor. [113:03] It is tough to follow Prabhaker. [113:04] It always is. [113:06] Until recently, I was the CEO of Apigee and have been now at [113:10] Google for 3 1/2 months and it's been great. [113:15] I want to do a special shout-out to a badass woman pioneer in the [113:20] tech space, Diane Greene. [113:22] It is great to work with her and for her and the opportunity to [113:26] continue to do it for quite some time to come. [113:32] Oh, she's actually in the audience. [113:34] Sorry, I didn't realize that. [113:38] It is awesome to see so many customers and partners here and [113:43] so many others that we will soon have an opportunity to [113:47] work with. [113:48] We thank you very much for giving us the opportunity to [113:51] serve you, and we hope to do that for many, many, many years [113:54] to come. [113:56] In addition to all the great innovation, all the great [113:59] innovation you've heard, one of the conversations that we have [114:03] with the G Suite for the companies that you work for and [114:07] board members is about how the Cloud enables innovative [114:11] business models, new ways of you creating and changing your [114:16] business that you wouldn't have been able to do before. [114:20] And we think a large part - a large part of creating [114:25] innovative business models is actually working on ecosystems. [114:30] In fact, Gartner actually in 2017 CIO agenda talked about how [114:37] some of the top performers of companies that they track [114:42] actually participate in many digital ecosystems. [114:47] Interestingly enough, not all ecosystems are the same. [114:51] There are many different kinds of ecosystems. [114:53] And generally when we think about ecosystems, we say, you [114:57] know, it must be public ecosystems. [114:59] But there are many different kinds and all of them affect [115:02] innovative business models. [115:04] So let's spend a couple of minutes talking about them. [115:08] T-Mobile is a great example - [115:10] great example of an internal ecosystem. [115:13] Cody Sanford, the CIO, has taken their core components like [115:17] billing, like customer, like product, and assigned senior [115:21] leaders to them. [115:22] These senior leaders are responsible for making sure that [115:28] these core components are available as APIs with SLAs; by [115:32] the way, only for the internal folks at T-Mobile. [115:36] In addition to making them available as APIs, they also [115:40] have to modernize the stack underneath the APIs. [115:44] And so what this is helping T-Mobile do is deliver products [115:48] and services at a pace that they have never done before. [115:51] We're going to see so much more from this team soon. [115:56] ABnet is one of the largest distributors of electronic [116:00] components and solutions. [116:03] Recently, they wanted to get into the Cloud [116:05] solutions business. [116:07] They set up a Cloud ecosystem. [116:10] They now have 5,000 resellers - 5,000 resellers. [116:15] And their Cloud business is growing over 1,000% - at a pace [116:19] of 1,000% year-over-year. [116:23] Industry ecosystems are actually well understood. [116:28] The stakeholders in healthcare IT have all come together. [116:32] These are the providers, these are the pairs, these are the [116:35] device makers, they are all coming together to create FHIR, [116:38] which is an interoperability standard to securely exchange [116:42] patient information. [116:43] FHIR is picking up momentum. [116:49] I love the Ticketmaster story. [116:52] Ticketmaster actually has a billion dollar API internally [116:54] and API that actually processes over a billion dollars on an [116:58] annual basis. [116:59] The aspirations are to become the operating system for live [117:02] entertainment worldwide. [117:05] They want to create a billion dollar business in a [117:07] public ecosystem. [117:08] They want all of you to be able to use Ticketmaster [117:12] functionality in the apps that you create; any apps, whether [117:16] it's in your car, whether it's your mobile device, or any other [117:19] kind of app that you create. [117:21] We think they are well on their way. [117:23] What's common across all these ecosystems is they are all [117:27] focused on one common goal. [117:30] That is to create innovative business models. [117:33] We think to create innovative business models you need three [117:37] different parts. [117:38] There are three pillars to creating innovative [117:40] business models. [117:41] You need to create and join ecosystems, you need to be able [117:46] to connect to apps, to data, and to devices internally and [117:51] externally, and be able to leverage all the great [117:55] innovation happening around you. [117:57] So let's spend a couple of minutes talking about each one [118:00] of these. [118:02] The question that all of you have to ask and every G Suite [118:05] has to ask, every board member has to ask, is your enterprise [118:11] ecosystem ready? [118:14] Because the right ecosystem platform can help your [118:17] enterprise with keeping an inventory of products or [118:21] solutions if that's what you do, apply pricing models to these [118:26] products, different kinds of pricing models, [118:28] consumption-based, many others, be able to work through multiple [118:33] tiers of partners because it's going to be B to B to C in some [118:36] cases, so multiple tiers of partners, and the most important [118:41] thing, to make your product very easy to discover and easy to buy [118:47] for end-user customers. [118:50] Orbitera is a multi-tier Cloud ecosystem that helps ISVs and [118:57] enterprises take their Cloud-ready solutions and help [119:02] them distribute and sell them. [119:04] We've had great momentum with Orbitera recently. [119:07] We are now over three billion transactions per month and the [119:10] growth continues. [119:13] APIs are well understood and everybody knows that it's the [119:17] cornerstone of every digital transformation happening in any [119:21] and every industry. [119:23] APIs come in different types. [119:24] There are APIs as a service. [119:26] This is how software talks to software. [119:29] And now with microservices, you're going to have thousands [119:32] of APIs in your enterprises. [119:34] And then there is APIs as interactions. [119:38] This is how your software talks to the physical world, whether [119:42] it's a mobile device, whether it's Google Home, or whether [119:45] it's your car, all through one API. [119:49] APIs as products is just getting understood. [119:53] Technology companies actually understand it well. [119:56] Now enterprises are starting to get it as well. [119:59] It is about taking a business focus to your API. [120:02] Thinking about your API as a business and bringing everything [120:07] that you can think about, product management, revenue, [120:11] channels, thinking about how the usage patterns are, thinking [120:15] about having a road map for the API that you publish, thinking [120:19] about all the different version and every engineering schedule [120:22] that goes with it. [120:24] So there is a lot happening with APIs as a product with some of [120:27] our enterprise customers as well. [120:29] Obviously an API platform needs to cater to the entire spectrum [120:34] of APIs. [120:35] And in addition, it needs to be very secure, it needs to scale, [120:40] it obviously needs to be multi-Cloud, it needs to work in [120:43] your private Cloud as well, and it needs to be able to support [120:47] an hybrid architecture. [120:52] At Apigee, we know a thing or two about APIs. [120:57] We process billions of API calls in our Cloud. [121:01] You process billions of API calls using our technology in [121:04] your Cloud. [121:06] Thank you very much for the customers that have bet on us. [121:08] Hopefully we get a chance to serve more of you soon. [121:13] Leveraging innovation is something that start-ups get [121:17] because they're born with constraints. [121:19] They start every day and say what is our core value? [121:22] What am I going to do that is going to be world class? [121:26] Everything else, they don't outsource but they go and [121:29] partner for. [121:31] It's something that happens really, really well in the [121:33] tech industry. [121:34] And now it's good to see enterprise companies do that [121:37] as well. [121:39] Audi, Crate and Barrel, and many others are taking maps APIs and [121:43] making it part of the customer experience that they are [121:46] delivering every day. [121:50] It's not just about the maps API. [121:52] There are thousands of APIs. [121:54] AccuWeather's weather API, Twilio's telephony API, and on [121:58] and on that are available that you should try to leverage to [122:01] accelerate your ecosystem. [122:05] As Google, we pioneered the product API space. [122:08] Google Maps has been around for over a decade. [122:12] It now has close to three million daily active users. [122:16] And you've probably heard in the last two days the number of APIs [122:19] that we are going to bring to you all with one simple goal is [122:23] to accelerate your journey as you think about more innovative [122:27] business models. [122:30] So we take these three pillars and bring them together and call [122:33] it the Connected Business Platform. [122:37] A Connected Business Platform is about creating and joining [122:40] ecosystems, internal and external ecosystems, all kinds [122:44] of ecosystems. [122:46] It's about connecting to apps, data, and devices in your data [122:51] centers, in your firewall, as well as externally. [122:56] And it's about leveraging all this great innovation that is [122:59] happening in the bazaar. [123:03] Digital is happening today. [123:07] You work for a company that has either been disrupted, is [123:11] getting disrupted, or is going to get disrupted. [123:14] It's going to happen with every industry. [123:17] Every analyst is writing about it, every strategy group is [123:20] writing about it, and more importantly, you are living it. [123:23] We think the time to act and to go off and think about creative [123:28] new business models is now. [123:31] As Google, we have 7 one billion user apps that we run every day. [123:39] We have created multiple ecosystems and have joined many, [123:43] many more. [123:44] We want to take our experience and meet your requirements for [123:49] today and partner with you for the future. [123:55] Please come by our showcase. [123:57] It's right here outside the third - on the third [123:59] floor itself. [124:01] Come and take a look at the great examples we have. [124:03] We have many enterprises that are building different kinds of [124:06] ecosystems and we are happy to talk to you more about helping [124:09] you with your journey. [124:14] Accenture has been a great partner. [124:16] We've participated in many Accenture tech vision reports. [124:23] The one in 2017 is phenomenal. [124:26] It talks a lot about ecosystems, AI, and many other things. [124:29] But I thought this quote was really interesting because it [124:32] talks about how companies need to partner with their customers, [124:37] a long-time pet peeve, the best product managers are your [124:39] customers and the employees, and this requires a cultural shift. [124:46] So to discuss this, I would like to invite Gene Reznik to [124:49] the stage. [124:58] Gene, take your time. [125:04] They already know we're going to take another two hours. [125:07] >> GENE REZNIK: Twenty minutes. [125:08] Twenty minutes. [125:09] >> CHET KAPOOR: That's all right we said, right? [125:10] Gene, thank you very much for joining us. [125:12] >> GENE REZNIK: Pleasure. [125:13] >> CHET KAPOOR: Would you introduce yourself? [125:16] Tell us a little bit about how Accenture thinks [125:19] about ecosystems. [125:20] And then most importantly, what are you seeing in the market? [125:25] What are enterprises thinking about and doing with ecosystems? [125:28] >> GENE REZNIK: Yeah. Yeah. [125:30] So I'm responsible for Accenture's ecosystem [125:33] and ventures. [125:34] What that really means to us - Chet, as you said - every [125:37] industry, every enterprise customer is being disrupted or [125:41] is doing a disruption. [125:43] What we really wanted to do is coming from the strength of [125:46] Accenture is really to help our clients on that journey. [125:50] And due to our partnership with Apigee that we started 5-7 years [125:53] ago, and now we're very excited to continue it as part of the [125:56] Google family. [125:58] And really what we see is this concept that you brought up, [126:00] externalize innovation. [126:02] To compete, you need to work much more broadly. [126:06] And fundamentally, innovation is the operative word with many of [126:09] our clients. [126:10] Now innovation is a culture, innovation is a set of business [126:14] processes, and I think what our clients are really demanding for [126:17] is for innovation to be also not hindered by integration. [126:21] I think this is where the technology and what technology [126:23] enables really needs to reinforce, the velocity, the [126:26] creativity, the business models that are really a lot of the [126:30] Fortune 100, 1,000, really want to set up to take their business [126:34] to the next level. [126:35] That's really the excitement for us. [126:36] We're really disrupting with them and helping them evolve and [126:40] really empower their businesses in the new. [126:43] >> CHET KAPOOR: That's awesome. [126:44] So the one thing that is different about the 2017 tech [126:49] vision report than the others that we participated in and work [126:52] with you on is that you come up with this concept called [126:55] people first. [126:56] >> GENE REZNIK: Right. [126:58] >> CHET KAPOOR: As we talked to the G Suite and about their [127:01] journeys, the cultural ramifications are [127:03] quite significant. [127:05] Can you tell us a little bit more about this people first [127:07] concept and how you're thinking about it and how you're talking [127:09] to your clients about it? [127:11] >> GENE REZNIK: Yeah. [127:12] I mean, over the past couple of years, I think we've all [127:14] realized that it's really about the people, the employees, the [127:18] innovation, the culture, but fundamentally it's probably the [127:21] hardest thing to transform. [127:23] I think a lot of us work to really reinvent our teams. [127:26] A lot of us work, even us consultants, to reinvent and be [127:28] relevant and ultimately embrace things like design thinking and [127:33] storytelling and a lot of the things that we saw here [127:36] earlier today. [127:37] And again, how did that really translate into you don't want to [127:40] come out of that and then go into the standard waterfall [127:43] development lifecycle that takes six months to do the integration [127:47] that you're trying to do. [127:48] You want it realtime. [127:49] You want to drag and drop. [127:51] You really want to continue and extend the culture that you're [127:54] fostering through your design studios, through your innovation [127:57] centers, all the way through the entire lifecycle and really [128:01] create those experiences for your customers and the ways that [128:04] you do your business. [128:05] So it's really an important part of what we're trying to [128:07] get right. [128:08] >> CHET KAPOOR: Awesome. [128:09] Thank you very much, Gene. [128:10] I look forward to working with you more. [128:11] >> GENE REZNIK: Absolutely. [128:11] Thank you, Chet. [128:12] Thank you. [128:19] >> CHET KAPOOR: Next I would like to invite three of [128:20] our customers. [128:22] These are customers that are actually already doing [128:25] ecosystems, a company that has actually been doing ecosystems [128:28] for quite some time. [128:29] They now are thinking about it differently. [128:32] And a couple others that are actually new to this and are [128:35] changing the way their companies work. [128:37] So please welcome our customer panel. [128:57] >> LYNN LUCAS: Hello. [128:58] >> CHET KAPOOR: You're ready to extend this for another [128:58] two hours? [129:00] >> LYNN LUCAS: I'm getting hungry. [129:01] >> CHET KAPOOR: All right. [129:03] So why don't each of you start? [129:05] And maybe I'll start with you, Lynn. [129:07] Introduce yourself. [129:10] Veritas and Google had an announcement yesterday. [129:13] Tell us a little bit about that. [129:15] And then tell us a little bit about how Veritas thinks about [129:18] ecosystems because you're not new to this. [129:20] I mean, you've been a tech company for a long time. [129:23] How are you thinking about ecosystems differently now? [129:25] >> LYNN LUCAS: Great. [129:26] So thank you very much. [129:27] Lynn Lucas. [129:28] I'm CMO and I lead marketing at Veritas. [129:31] We made a major announcement and a strategic partnership with [129:34] Google yesterday. [129:36] It's really all about what's been talked about here, which is [129:39] the importance of data and how it's changing businesses. [129:43] How do you have visibility into your data? [129:45] How are you moving the right data to the Google Cloud? [129:49] How do you protect it there? [129:51] And then for you G Suite users, how are you ensuring that you [129:54] continue to have regulatory compliance to [129:58] increasing regulation? [129:59] Now, Chet, what you said is that Veritas has long understood the [130:04] importance of ecosystems. [130:06] We've invested heavily in it for years. [130:08] I mean, IT is built on the fact that you have to have ecosystems [130:13] to make it easier for you. [130:14] This couldn't be more important in the era of the Cloud. [130:17] When your data is spread amongst Google Cloud, but as Eric said, [130:22] probably more than just that. [130:24] Many of you are probably having your CRM, your HR applications [130:28] and other workloads in Clouds. [130:30] We're building ecosystems with Google Cloud and with many [130:33] others to ensure that it's easier for you to manage and [130:37] protect that data. [130:38] >> CHET KAPOOR: Awesome. [130:39] Thanks, Lynn. [130:40] Roger, same for you. [130:42] Give us an introduction. [130:43] And tell us, what is this Pitney Bowes, you know, been around for [130:48] 100 years. [130:49] What are you guys doing with ecosystems? [130:50] >> ROGER PILC: Yeah, absolutely. [130:51] So, Chief Innovation Officer of Pitney Bowes. [130:54] We are a global technology company that powers commerce. [130:57] We have been around for 100 years and we've been undergoing [131:00] a very, very exciting transformation. [131:03] That transformation has been powered by what we call the [131:07] Pitney Bowes Commerce Cloud Apigee. [131:09] As Chet knows, they've been an absolutely outstanding partner [131:11] in helping us digitize everything we do for clients as [131:15] we discovery, identify, locate, communicate, ship, and pay. [131:19] So the Commerce Cloud in Apigee and APIs has allowed us to power [131:22] a very exciting commerce ecosystem with many millions of [131:27] sellers, many millions of buyers, multiple e-commerce [131:30] marketplaces, tens of sellers, financial services companies, [131:34] and it's really helped transform our company. [131:37] More recently, we also engaged with Google around the Android [131:40] operating system for S&B appliance and our S&B ecosystem [131:44] and recently chose Orbitera as well as the apps to our [131:47] technology and our Pitney Bowes Commerce Cloud for S&B. [131:50] So it's been a great journey. [131:51] >> CHET KAPOOR: Awesome. [131:53] Fatala, I was really moved by Fred, your CEO's letter in the [132:00] annual report where he talked about we're going to stop being [132:04] a retailer with a digital area. [132:06] We're going to become a digital company with some physical space [132:10] and a human touch. [132:12] >> ANDRE FATALA: Right. [132:13] >> CHET KAPOOR: You've been doing this for a while. [132:14] There were two people and a dog in a small room when you first [132:18] started out. [132:20] Can you tell us - introduce yourself and tell us a little [132:22] bit about your journey as you've expanded Magazine Louiza's [132:28] perspective on how to think about ecosystems. [132:30] >> ANDRE FATALA: Okay, right. [132:31] So my name is Andre Fatala. [132:32] I am from Brazil. [132:34] So I am glad to be here talking about technology and [132:37] not carnival. [132:41] We started our journey in 2012. [132:44] We were a really small R&D team that tried cracking the systems [132:51] of the corporate enterprises company. [132:55] After two or three years, we just get all the points of sales [133:01] development inside the company using a lot of [133:03] lean methodologies. [133:06] And we moved everything to the Cloud since then. [133:11] We are starting for the APIs. [133:14] So when we get together and talk about how to leverage the [133:19] innovation and then decouple the big legacy systems. [133:25] Since then, we got this idea of our CEO to try to be a digital [133:33] platform with physical stores. [133:37] Since then, we just turned out entire company to do this thing. [133:42] We are seeing really good results. [133:45] Last year our digital just goes up 35%. [133:51] I don't know if you guys know, but in Brazil we are facing a [133:55] really bad economic crisis. [133:58] >> CHET KAPOOR: For sure. [133:58] That was great. [134:00] Fatala, one follow-up question for you. [134:02] I'm sure this was not easy, right? [134:04] I mean, yeah, you've started out in a small office and a dog and [134:09] everything and you've transformed quite a bit. [134:13] What was the hardest part of the journey? [134:16] >> ANDRE FATALA: I think the hardest part of the journey was [134:18] to get everybody understanding this changing of the mindset of [134:23] the company. [134:25] After that, we have the challenge to change the entire [134:29] technology to provide the flexibility and agility to to [134:34] the business. [134:35] But I think that we are doing a great job on that. [134:39] >> CHET KAPOOR: And so - go ahead. [134:40] >> LYNN LUCAS: If I could add into that, I think that the [134:43] transformations - and Veritas is undergoing that transformation [134:46] as well - the hardest part is the people and the culture. [134:50] It seems like that's what you've experienced as well. [134:53] The ability to move forward with the people and culture makes all [134:57] the technology come to life. [134:59] >> ANDRE FATALA: That's right. [134:59] Technology is there. [135:01] It can be used for everything. [135:02] We need to get the right people to do this. [135:06] >> CHET KAPOOR: So, Roger, how do you - I will ask a very [135:11] direct question - how do you deal with antibodies? [135:15] Right? I mean, it's a cultural shift. [135:16] It's understood in the G Suite, the board, everybody, we're [135:19] going to make the transformation, but not [135:22] everybody gets it. [135:23] How do you deal with those cultural aspects of [135:25] this transformation? [135:26] >> ROGER PILC: Yeah. [135:27] It's similar to the other panelists. [135:29] For us, culture has been the biggest positive in [135:31] our transformation. [135:33] We've had a company that for 100 years has valued our employees, [135:37] has had great employee engagement, has had a culture of [135:40] innovation, and most importantly collaboration. [135:44] For us, we were able to harness that. [135:47] That was a great surprise being there are only three companies. [135:50] So the key to us has been have a very clear vision of where we're [135:53] going, be very consistent with it, have a focus on operational [135:57] execution and the discipline of getting things done, and just [136:00] maintain a belief. [136:01] There are pockets of concerns where maybe there was past [136:06] failings, so maintaining that vision and that constant belief [136:09] and sharing that sense of belief has been critical. [136:12] >> CHET KAPOOR: And the sense that it is okay to move the [136:14] goalpost a bit. [136:16] >> ROGER PILC: Right. [136:16] Absolutely. [136:17] We move them very far. [136:20] >> CHET KAPOOR: Final question for all three of you. [136:22] We'll start with you, Fatala. [136:26] If you were in the audience a year ago, what advice would you [136:28] have liked to have given yourself? [136:31] >> ANDRE FATALA: Try to focus on our customers, be creative to [136:36] solve their problems and not ours, and I think when you [136:42] decide to do something, execute like crazy. [136:46] You need to try a lot. [136:48] Get data, get some insight from this data, and then try to [136:53] evolve our product. [136:55] >> CHET KAPOOR: Awesome. [136:55] How about you, Roger? [136:58] >> ROGER PILC: The key is to believe that great things and [137:01] great transformation is possible. [137:04] With the right vision and recognizing the availability of [137:07] great technologies today, picking the absolute right [137:10] technology partners, we couldn't misstep on that. [137:13] So choosing the right technology partners and executing against a [137:16] vision has been critical. [137:17] >> CHET KAPOOR: I love that. [137:18] You're starting with believing. [137:20] >> ROGER PILC: Right. [137:21] >> CHET KAPOOR: And Lynn, how about you? [137:23] >> LYNN LUCAS: If it's your digital transformation, you [137:25] absolutely have to focus on people and culture first. [137:29] When it comes to data, which is the heart of how we build our [137:32] businesses, get visibility into what you have, 30% of it is junk [137:37] or rot, and then move it quickly to the Google Cloud, rethink [137:42] your storage, reduce waste, make a more sustainable environment. [137:46] >> CHET KAPOOR: Awesome. [137:47] Thank you very much. [137:49] I would like to now invite Urs back on stage. [137:58] >> URS HOLZLE: All right. [137:59] So you see a lot of innovation at this conference behind me. [138:02] You see actually just a subset of what we announced today [138:06] because we don't really - we can't really fit anything [138:09] on one slide. [138:10] We showed how Google Cloud has the best infrastructure, the [138:14] best security, the best productivity suite, and we [138:17] showed a host of new collaboration features, [138:20] including the awesome Jamboard. [138:22] So I'm really looking forward to that. [138:24] And then, of course, there is a lot of help to get you started [138:27] where you are, and we also showed you yesterday how we're [138:30] building our ecosystem with global partnerships with [138:33] companies like Pivotal, SAP, Rackspace. [138:37] And we understand it's not just about the technology, it's about [138:40] helping everyone be successful in using it. [138:44] So that's what this conference is all about. [138:46] So thank you very much for coming, and thank you for the [138:49] trust that you've placed in us. [138:51] Have a great day.