>> DIANE GREENE: Thank
you so much and welcome. This is just an incredibly
exciting event for all of us in Google Cloud, for all of
us at Google I would say. You know, the guy who was micing
me was telling me that there was a queue around the block, that
we had a blockbuster, and I'm sorry for the queue but you
know, we're pretty happy to have all of you here to be at
capacity at 10,000, all of the people livestreaming. I really regret the live
streamers couldn't see the surround on that amazing
animation that we just showed. You know, Google -
I mean, Cloud, Cloud. Does anybody not agree it's the
biggest thing going on in IT right now? Yeah. Woo. And Google Cloud, I mean, the
adoption of cloud in general, I believe it must be accelerating. I know at Google
Cloud this year we've seen unbelievable acceleration. What's that? Okay. Okay. And the quality of
our customer conversations are really changing. We - you know, for a while it
was more let's, let's use big query. Let's do our data analytics. Let's do our
machine learning project. But just two weeks ago I met
with five customers over the course of the week and three of
them said hey, I want to do a full lift and shift. I just want to move
everything to the cloud. And it was only last September
that we actually started calling ourselves Google Cloud, all that
we do from the infrastructure to the workplace collaboration,
productivity, and even the mobile devices
in the enterprise. It's been a remarkable year. I mean, our engineers have done
over five hundred releases. These are releases to get
functionality to our customers. These are new innovations. These are inventions and just
ongoing improvement in all that we do. Security,
reliability, efficiencies. Cloud is just an incredible cool
place to be working right now. It's basically revolutionizing -
it's sort of where a lot of the revolution, the digital
revolution for every industry is going on. In financial services, you know,
tremendous efficiencies on how data is processed or how the
banks can talk directly to their customers. In health, we're seeing a
revolution in diagnosis and predicting patient outcomes. In retail, particularly exciting
area, taking billions of dollars out as we revolutionize
the store of the future. Media, something Google is
particularly well situated to do, and energy, manufacturing,
optimizing, reducing the energy usage, making the factories, the
manufacturing more efficient. It's incredibly
exciting to be part of this. We have such
interesting customers. We have a customer Planet Labs
that is taking pictures of the entire earth every three
hours and that is going on Google Cloud. I think the cloud, you know, it
is no longer sort of a utility for surplus, peak capacity. It is no longer a
place just to store things. You know, startups, it's where
they start, but it's not solely for a startup to figure out what
- what demand they are going to have before
they spend any money. Workplace productivity has
really moved beyond sending files around via email. And the cloud is really
what gives - what's giving - increasingly giving customers
their competitive advantage. And what is important there,
paramount is security, from the board level on down. Nobody - you know, everybody
really cares about security and I will say a little more
about security at scale in a little bit. It's about cost, performance
effectivity, the flexible costs, the lock-in,
and it's about reliability. And I am just, because of all
of the interest in reliability recently, I am just
going to say a few words. You know, you
take Google Search. It was designed. It runs as you see
when you do a search. It runs at
five-nines of availability. That is 99.999% uptime. And we're - our cloud, that is
how we designed our cloud and we're making it really easy
for our customers to design and deploy for that kind of
reliability, distributed, no single point of failure. And I'm really proud. I just learned yesterday that
we were recognized as being - having the highest availability
of any cloud over the course of 2016. Thank you. I think 2017 will
be promising too. We're putting a lot of effort
across the board, across our entire systems with the
unbelievably excellent engineers we have to keep
giving you more nines. What else? Well the world customers are
getting increasingly focused on data. You want the world's best data
analytics, the world's best machine learning. This is definitely
a Google strength. And then finally
workplace culture. You know, really it's about
real-time collaboration. It's about finding ways to get
the bureaucracy out of a company because everybody has got to
move fast now and we have the technology tools to do that. So over the next three days,
you're going to hear from our engineers. They're going to tell
you about our products. They're going to
give you our roadmaps. You're going to
see lots of demos. Our customer facing people will
help you navigate Google Cloud. And you're also, in this
keynote, going to hear from our customers. Everybody at Google Cloud is
so incredibly proud of the customers that we have here
today that we've been working with over the
course of the year. Six customers
and a major partner. So now I just want to pause. I am wearing my red ribbon. I want to acknowledge that it is
International Women's Day today. Wow. I have to say that
response is pretty exciting. You know, this industry - I
think I have been incredibly fortunate but it was sort of an
industry where I - you know, I was lucky and I kind of was -
chose to be kind of oblivious to what was going on, and now we're
in an environment where women are just increasingly having a
huge impact and adding a lot of value to our industry. And women are celebrated. If they raise their hand, they
say hey, you're missing my value; you're not
recognizing what I'm doing. And at Google, we strive
at Google Cloud to have an environment where no one needs
to raise their hand, but no matter what it's completely
safe to do that, okay? You know, I really look forward
to the day when this audience is maybe 50% women. It's more fun to have diversity. So shifting gears, now
I want to bring out our next keynote speaker. It is my phenomenal privilege to
introduce our incredibly smart, compassionate, and just
great leader, Sundar Pichai. >> SUNDAR PICHAI: Thank you. This is a big crowd. I have done many I/O
presentations here and it feels like a bigger crowd
than Google I/O. So thank you all
for joining today. It's great to be here. Google was founded on a
mission to organize the world's information and make it
universally accessible and useful. Over many years, we've had the
privilege to create products that have served billions of
consumers across the world. To do that, we've had to engage
pretty much every branch of computer science and we have
had breakthroughs along the way. It had led us to breakthroughs
in cloud computing technology, collaboration software,
analytics, things which are now transforming
businesses everywhere. So with Google Cloud, now we
offer businesses Google quality computing, security, and
business tools, and it is an extraordinary big bet for us
both as a platform internally for our own innovation but
also as a way of bringing our infrastructure to serve the
needs of businesses across the world. Our customers and partners will
see many ways to leverage all of Google, be it Google Maps,
Android, YouTube, our devices ecosystem, and will deliver it
over the most efficient, high performing, and secure
core computing resources. For over fifteen years, we have
also been investing in machine learning and AI and as
businesses increasingly shift online and create more
information, they can leverage these AI and machine learning
tools to understand the information and build
better products that delight more customers. Above all, we are excited to see
what you all will create with our products and services. When we shift things on the
consumer side, we always get surprised in how they use
our technology in a thousand unexpected ways. It makes us better as a company. With Google Cloud, we hope
to have the same journey. We hope to grow and
learn alongside you. Partners already have pushed us
to add more ways to collaborate beyond video conferencing. They have exposed us to more
real-world scaling needs, for example the demand that
retailers like Home Depot get on Black Friday. They have shown us exciting new
use cases for Google products. You will see what EBay has done
with Google Home as an example. In many ways to me, Google Cloud
is a natural extension of our mission to make information
accessible and useful. We're just doing
it for businesses. It means increasing the
availability and lowering the cost of data storage
and computing resources. It means making it easier for
businesses to understand and act on critical information. That is what G Suite does. G Suite enhances information
sharing and collaboration. Machine learning finds
previously hidden patterns in data, new information which
helps businesses improve their products and better
serve their customers. So I am really excited by the
progress here and I can't wait to see what you all build next. So without further delay, I am
going to invite Diane back onto the stage. She joined us in late 2015
and the business has grown dramatically for us
under her leadership. And given that it's
International Women's Day, it is also a pleasure to welcome back
onstage, a trailblazer for women in technology. So let's have Diane back to
share more about how Google Cloud is helping all of you. Thank you. >> DIANE GREENE: Thank you. That was really great. Thank you so much, Sundar. We really are one Google. And at Cloud, you know, one
of the reasons I joined; we're leveraging almost two decades of
innovation and technology really built for the kind of
enterprises we have today or we're starting to have. I want to just kind of
underscore or say a few words about our infrastructure. We don't really normally give
out many numbers and I'm not going to necessarily break that
policy but I am going to tell you some
interesting nuggets, okay? We have possibility the world's
biggest, most powerful secured network, okay? Tomorrow, Urs Hölzle, who really
is behind - employee number eight behind our data centers
and infrastructure is going to go into this more and tell you
about what we're doing and where we're going. But today what I am going to
tell you about is just the kind of the scale of
our infrastructure. You know, if you took not even
our largest data center and stacked all of the servers
one after another, it would be higher than more than five
thousand feet above Mount Everest, okay? I was just going to say above
Mount Everest but the person that calculated this said no,
no, add the five thousand feet. So - just one building in our
South Carolina campus is bigger than Stanford's
Arrillaga Football Stadium. We have a new
complex in the Netherlands. It's got over ten
thousand miles of cable. That is enough to go from
here to Tokyo and back. And then one of the things we're
all just very proud of at Google is we've been focused on
being carbon neutral from the beginning. You know, compute
takes a lot of power. Someday we'll be as efficient
as our brains but it does take a lot of power. Google is the largest
purchaser of renewable energy. That's wind farms. That's solar. In fact, our purchases are
larger than the next three or four purchasers combined. So why did I tell
you about that scale? Well I want to give you some
examples of why it matters, just to customers. You couldn't have failed
to notice that Niantic Labs launched Pokémon Go earlier this
year and just six days after they released, they kind of
eclipsed their most optimistic three month predictions. Their queries
increased fifty fold. So they were running on Google
Cloud and our site reliability engineers were with them
twenty-four hours a day, seven days a week making
this thing go flawlessly. They were there
around the clock. In fact, it's when we realized
we needed the same kind of functionality for our customers. Pokémon Go - last week at the
World Congress, they announced - Niantic announced that they have
had 88 billion Pokémon caught. Painless scaling
on Google Cloud. The other example is an
enterprise customer, very innovative retailer, Home Depot. They are now running on Google
Cloud and we got them through Black Friday. It was the most painless Black
Friday ever and they will be here later today
to talk about it. Also with that scale comes the
need for phenomenal security and I am just going to go
into that a little. This shows sort of the layers of
security and you can see there are many gaps in between the
networks, the servers, and the software. So we built our own
chips in the network, in the computing devices. We have to look at
every single end point. We actually have over seven
hundred people just fulltime worrying about security,
watching the network, securing Google Cloud. And if you add a Chromebook
which this list Chrome, you know, doesn't store, you get
an added layer of security end point. I personally only
use a Chromebook. By the way, Chromebooks last
year outsold Mac OS in the - Macs in the United States. And something we're really
proud of is what we're doing for education. Over 20 million Chromebooks used
by students in schools worldwide and growing. Yeah. Thank you. So cloud is just a
transformational technology. You know, it's changing
how people architect their information, use their data. It's changing how they work with
their customers, how they work together, you know. In effect, our companies
are becoming virtualized. We're operating both in the
physical world and the virtual and there are many things you
can do in the virtual world more powerfully than you can in the
physical, particularly as we all start taking advantage
of machine learning. You know, I really believe
that the cloud with the best technology is the best cloud. It gives you a competitive edge. And you want to be on a cloud
that is going to keep you on the edge of, you know - technology
is moving so fast and you want to maintain that
competitive advantage. But all of this technology
needs phenomenal customer focus. We've really doubled down on our
customer facing part of Google and I am super - as I said,
super proud today to give you those proof points to have our
customers here to tell you about that, customers and partners. Just to give you a preview,
we've got Disney here. We've got our partner SAP. We've got Colgate-Palmolive,
we've got Verizon. We've got Home Depot and we've
got HSBC, one of the world's largest banks, and then we'll
close out with EBay with a pretty wonderful demo. So I'm going to begin by
bringing out one of the most loved brands in the industry. Disney was here last year and
they were just dabbling in the cloud, just starting to do some
things, and they actually got more done than they expected. They are ahead of schedule. They are all in. Life and shift. And then they have aspirations
for even bolder things we will do together. So I thought it would be very
interesting to start out with the CTO and SVP of Disney,
Consumer Products and Media - Interactive Media. Mike White. Please welcome him to the stage. Thanks so much. Okay. Yeah, we have chairs. I was worried. I was wondering where they were. >> MIKE WHITE: Yeah. >> DIANE GREENE: It's so
great to have you back, Mike. Thanks so much. >> MIKE WHITE: Thank you. >> DIANE GREENE: You know,
we just started talking about moving Disney Interactive
Consumer Products to the cloud and now you're taking
a cloud first approach. Can we get an update? >> MIKE WHITE: Yeah, for sure. First off, we really looked at
moving to the cloud as a way to transform the way we
developed software. I can tell you it's been truly
transformative for us from software
development perspective. Currently, we have about five
hundred projects in the cloud. That includes everything from
customer facing or guest facing products and production we call
it, to dev, to QA, to our R&D environments, and also all of
our data suites are running on a big query. We've started with a cloud
first approach for all new applications since the last time
we talked and we've been - had the luxury actually of being
able to re-architect a lot of our legacy systems for the cloud
and as you mentioned, we're ahead of schedule. That's really a testament
to - you said the word partnership earlier. It truly is a partnership that
we have developed with Google that has been, you know,
really helpful for us. So thank you for that. >> DIANE GREENE:
Oh, you're welcome. I think our teams just
really clicked and really just collaborated phenomenally to
bring your vision to life. >> MIKE WHITE: Yeah, for sure. I mean, I think - you know,
it was really great talking developers to developers and
making developers come in first for this. But what was really interesting
was within our division, we have different application
characteristics. So we have some
really top ranking games. We have kids chat app. We have several Disney sites
that are all running out of Google Cloud and that is
both internationally and domestically. But what that's really done for
us, and I talked a little bit about being transformative from
the software development side, is it has really been able to
free up our developers and our resources. The way we look at our
developers is they're, you know, technologists that
are storytellers. For us, we use technology
in service of the story. That's what we do at Disney. It's been our legacy since Walt. So we've really been able to
focus on telling great stories and with that, we've been
really taking a machine learning mindset as we're
on our road to AI. So if you think about character
development, extension of characters, and really getting
closer to the characters, that is where we want to play. And really with this first lens
of machine learning I think in the short-term, we're really
looking at how that will impact retail which we're
really excited about. >> DIANE GREENE: It sounds
like magic helping magic. Do you mean just for your
operations you're doing this or is it really
touching the customer? >> MIKE WHITE: It is both. So for sure, we're leveraging
things like image recognition technology for some backend
services, but we also have 8,000 characters within Marvel and you
can imagine all of the images that come with that. We're also looking at ways to
really strengthen the connection of our consumers through
optimization, personalization, and really enhanced experiences. So those are customer
facing and backend as well. We look at that as kind of
a symbiotic relationship. >> DIANE GREENE: You know, I
just love your mission at Disney to bring this magic, you know,
to the daily lives of people and - and your fans
around the world. It is so great to support that. >> MIKE WHITE: Yeah, I mean,
that's what gets us up in the morning, right? I mean, our company's stories
and characters mean so much to so many people. I mean, for everyone in the
audience I can imagine that you have a favorite character that
is Disney or Pixar or Marvel or Star Wars and with that,
maybe you have multiple which is great. So that's really what we do. We use technology to empower
and really create the magic and we're really excited with the
potential of machine learning and AI, what that can do for us. So thank you. >> DIANE GREENE:
Thank you, Mike. You know, I can't wait to
see where we are next year. Thanks so much. >> MIKE WHITE: Thanks so much. >> DIANE GREENE:
Great having you here. >> MIKE WHITE: Thank you. >> DIANE GREENE: You know, it's
pretty exciting for us to see this full lift and shift
happening from just sort of an initial fairly small engagement,
sort of what I was describing earlier, and then next year
doing even cooler things with our commitment to really
just all out partnering with our customers. Now you know - what actually -
I just want to mention that in 1955, the average
age of a company was about seventy-five years. Now in this year, the average
age of a company is only about seventeen years. But all of our customers and
partners that are here today have been around much longer
than that and are beating the odds, even EBay, you
know, at twenty-two years. I don't mean even EBay. I mean even our
youngest has been longer. I love to partner. I love to work with a community
of technology companies. And you know, we can
build on their offerings. They can build on ours and it is
just terrific for the customers. Customers first always,
then our partners. We've been working with SAP for
over, you know, over the course of the last year and the more we
have been working together, the more we see what we can do. As recently in Germany it was
very exciting to be, you know, at this, you know, the world's
biggest ERP firm, a company that when I started working in the IT
industry at SciBase pre-public, they were already, you know,
such an impressive company. So today I am really thrilled
to welcome to the stage Google Cloud's latest major strategic
partner Bernd Leukert, Head of Technology and Innovation and
Member of the Executive Board at SAP. Hi, Bernd. >> BERND LEUKERT:
Thank you, Diane. >> DIANE GREENE: Oh, thank you. >> BERND LEUKERT: And a warm
welcome to the Google Cloud Attendees as well from my side. >> DIANE GREENE: Yeah. >> BERND LEUKERT: I mean,
you said it already. Representing here SAP as a
company which now for almost forty-five years is enabling
companies to succeed across twenty-five industries, across
eleven lines of business with our cloud and in
historical times as well, our on premise products. >> DIANE GREENE: Yes. >> BERND LEUKERT: And I mean,
you refer to the former times. I still remember joining SAP
when mainframe computing was the flagship. We had R2 at that point in time. And since that time, we have led
many technology waves and helped our customers to
succeed going forward. And as you saw, that technology
passion - I liked what you said - and our focus on customers has
rewarded us, that we are able to serve now more than
345,000 customers. I have to say
enterprise customers. Creating products and used by
millions of people and decision makers day in and day out. And one of our most important
innovations just recently is of course a growth engine for us. The last couple of days. This is our flagship
product, SAP HANA. SAP HANA is our in memory data
and analytics platform and in the meantime as well, used by
more than 14,000 customers. And these customers implement
scenarios that leverage HANA's advanced analytical processing. They use spatial crafts,
streaming data, text analytics, as well as
predictive capabilities. And to turn machine learning and
insight into real intelligence and then follow it into action. Of course as well, use the
platform for our traditional transaction business. >> DIANE GREENE: You know, it's
just such an impressive story. Hasso Plattner and you and how
you've been so agile and keep reinventing SAP and you know, I
first was aware of Hasso because of my sailboat racing. He has very good taste in boats
but he's also been great for design thinking you've led in
the mainframe, in the client server, and now making a core
part of your business the cloud. >> BERND LEUKERT: And I mean,
Hasso is basically one of our founders and still the
technology advisor and we have now as SAP, the goal these days
to become the cloud company powered by HANA. But now we are ready to
take things a step further. And today we are delighted to
announce here in San Francisco, the general availability of SAP
HANA on Google Cloud platform. >> DIANE GREENE: Yes. >> BERND LEUKERT: Of course -
of course if anybody questions, with full support from our SAP's
existing support contracts. No doubt about that. And we will offer enterprises
the world's leading best network, the fastest service,
and the best enterprise data platform for mission critical
applications, analytics in one stop. And for the developers and the
engineers in the audience, I am also pleased to announce that
the developer edition for SAP HANA - we call it HANA Express -
is also being offered in Google Cloud Launcher which is Google's
marketplace for enterprise-grade partner applications. And with just a laptop, it is
going to be easy to build SAP software application capable
of instantly working with the largest possible suite of
products, accessing data primarily in the cloud world. And obviously, this is great
news for our corporate software developers, and I mean our joint
corporate software developers, looking to build new products on
existing data and applications. >> DIANE GREENE: Thank you. >> BERND LEUKERT: Our cloud
partnership story with you, with Google, does not stop here. This is just the starting point. >> DIANE GREENE: That's right. >> BERND LEUKERT: We are
working on running the SAP Cloud Platform, and specifically our
cloud foundry based platform as a service framework with a host
of business and integration services as well on the
Google Cloud Platform. And we recognize Google
definitely is a frontrunner, is a frontrunner in technologies
and specifically in topics like containerization, in scalable
data processing which are featured prominently here at
the Google Next Conference. And our SAP Cloud Platform and
big data teams will work very closely on with your teams - >> DIANE GREENE: Yes. >> BERND LEUKERT: On making
technology such as Kubernetes, such as spanner
workforce scaling and enterprise workloads. And I see a tremendous
opportunity moving forward from that point for our
customers to build these scalable
enterprise applications. Using these endless combinations
of services, of APIs exposed on the Google Cloud Platform but as
well on the SAP Cloud Platform, and we will have that
offered together for you, for our customers. >> DIANE GREENE: Yes. >> BERND LEUKERT: And lastly,
our joint memorandum of understanding on machine
learning co-development is just the starting point. It's just the starting point and
you can expect to hear more from the two of us, from our
companies, and from our next big milestone is the SAP Sapphire
Event which is SAP's Google Next in May in Orlando. >> DIANE GREENE: You know, this
is just a tremendously important partnership to us here at Google
and - and our commitment really is to SAP and your
thousands of customers. As another - yet another but a
really big deal element of our partnership with SAP and Google
- between SAP and Google Cloud, we're collaborating to develop
leading industry governance, risk, and compliance. We want a solution
for the public cloud. And what that means is it's
going to enable SAP to be a data custodian of the customer's
data that is stored in GCP. It is increasingly important for
enterprises, you know, customers moving from on prem to the cloud
to retain their insight, their control while still benefitting,
you know, from the agility, the scale, from the
presence of a public cloud. And we see this joint solution
as the basis for a new way to approach compliance
and governance in public cloud environments. >> BERND LEUKERT: I completely
agree, Diane, and being with Google Cloud ensures from my
perspective, deployment to the greatest number of markets
and enterprise and visions. And our strategic intent is to
work on extending GCP's robust set of existing capabilities
through jointly developed solutions that bring global
visibility, access control which our customers ask for - >> DIANE GREENE: That's right. >> BERND LEUKERT:
To our customers. >> DIANE GREENE: Yes. >> BERND LEUKERT: For improved
risk and compliance management. And to serve as a data
custodian, we feel honored. And this is a first step. We analyze GCP's audit logs to
assure appropriate visibility and control, but again not for
us, again for our customers and ultimately work towards
improving access control requests and even - you
mentioned security before - encryption key management on
service provider access to the customer's data. >> DIANE GREENE: You know,
I really think that this partnership is going to improve
compliance and risk and you know, really ease the hassle
that thousands of companies - millions of
customers are facing. You know, we really see
opportunities also for more of the machine learning on SAP's
data and applications, getting our - you know, the enterprise
is just more value and insight from the data they already have,
you know, building these new applications. Yeah. This is - and that's why we're announcing these plans to work jointly on
our new machine learning. We're also really interested
in tighter integrations with G Suite. Is this your part? >> BERND LEUKERT: No. >> DIANE GREENE: Okay. With G Suite's
collaborative tools - anyhow. G Suite I will just say is a
platform also with open APIs and so we are integrating G Suite
with the SAP applications. I think that is going to be
super exciting for people, you know, really increase
customer satisfaction. And just to underscore that
we have a joint customer, just wonderful person, and - and from
a company that is actually 211 years old, I am really pleased
to welcome Mike Crowe, the Chief Information Officer from
Colgate-Palmolive if he can join us on stage. >> BERND LEUKERT: By the way,
a long-lasting friend of mine. >> DIANE GREENE: Hi, Mike. >> MIKE CROWE: Good morning. >> DIANE GREENE: Thank you. >> BERND LEUKERT: Mike,
great to see you again. >> DIANE GREENE: Great
to have you here, Mike. Please, why don't you just
tell us a little bit about your Colgate-Palmolive and your
partnerships with both of us? >> MIKE CROWE: Sure. So first of all,
thanks for having me. I'm happy to be here. Colgate is a leading global
consumer products company selling our products in 223
countries around the world. Our 38,000 Colgate people focus
on four core categories, oral care, personal care,
home care, and pet nutrition. Likewise, we have a very focused
IT strategy which at its core relies on a small number
of strategic partnerships. We chose SAP twenty-three years
ago and Google Cloud just last year for very
much the same reasons. You're both highly innovative
and you both have a willingness to partner with us. And we see tremendous value from
implementing your technology. We have of course been
innovating with SAP for years on business processes and we're
already starting to see the benefit of Google's
collaboration suite of applications. >> DIANE GREENE: So Mike, I
mean, you were one of the first potential customers I met when
I joined Google and now you've implemented G Suite. Can you tell us a
little bit about it? >> MIKE CROWE: Sure. So when we started our project
in May of last year and we were very encouraged by the positive
email responses we received from our employees as we announced
that we were going Google. Still, we knew that it would be
a major change management effort to go to a new
collaboration platform. So Diane, we worked with
your team as well as our implementation partner, SADA
Systems to lay out a phased approach for implementation. And in just three months,
we were able to launch for our global IT organization
all around the world. We learned from that and then we
leveraged that excitement that our employees had shown
to recruit a few thousand volunteers to be early adopters. And within that set, we had
about nine hundred Google Guides who were to be the champions on
the ground as we implemented and to be the
trainers as we rolled out. And all of that set us up
very nicely for our final implementation which was in one
weekend in November we took more than 20,000 users live. So all told, more than
28,000 users in six months. >> DIANE GREENE: Yeah, Mike. I mean, it's just phenomenal to
successfully move 23,000 people over the weekend and I can't
tell you how much I enjoyed the little video you sent me to
educate your company where you're sitting there talking
to Mr. C, a cartoon character. What have you seen
since you launched? >> MIKE CROWE: Well in just a
little over three months, we see that people are
working differently. Sure everybody is using mail and
calendar, but the usage of the other tools is quite impressive. Ninety percent of our users are
active on Google Drive and in February alone, we had more than
57,000 hours of video hangouts. You know, allowing our user
base to connect more easily both while in the office as well as
on mobile devices while outside the office. And this is critically important
for us as our Colgate teams around the world work
more and more together from different locations. And then finally, seeing
a faster uptake in the productivity suites in Docs
and Sheets and Slides as well, faster than I had expected. >> BERND LEUKERT: So Mike, I
still remember when you asked me last summer how is your
relationship to Google? And I was thinking, okay. We have a conversation. Now he is asking me. He knows me now from more than
two decades but now we are here together and
it's pretty amazing. But I have to ask you; now
we have made the first step. What excites you going forward? What should the two of us
deliver to you going forward? >> MIKE CROWE: First, I am very
excited about the Google SAP partnership, about the
announcements that have been made here today, and about the
possibilities that the combined strengths of these two
great companies bring out. I think it makes perfect sense. It's a perfect fit to combine
SAP's business - enterprise business application expertise
with Google's expertise on infrastructure as a service,
running in the Google Cloud Platform. Bernd, you and I have been
talking about incorporating machine learning into a broad
array of the SAP business processes for some time - >> BERND LEUKERT: Yes. >> MIKE CROWE: So I am
excited about that as well. And then one more example. As you know, we are users of
SAP's digital boardroom, a co-innovation project between
us and SAP which allows us to review the state of the business
at any time using the latest data, and we are already working
on incorporating Google Slides into that application as well. So in summary, I think the
possibilities are going to be numerous for us. >> DIANE GREENE: This has
been really terrific for us. We really enjoyed
working with both of you. It's great for Google Cloud. It's great for you guys. And we're really looking forward
to see what we're all going to do next. Thank you so much both of you. >> BERND LEUKERT: Thanks
for having us here, Diane. >> DIANE GREENE: Yeah, alright. >> BERND LEUKERT: Thanks a lot. >> DIANE GREENE: Thanks guys. That was great. >> MIKE CROWE: Thanks a lot. >> DIANE GREENE: So as Mike
was saying, G Suite is this fantastic collaboration suite. It is really transforming the
culture and the productivity at Colgate. One thing also that I will just
add about G Suite that we hear in every customer that deploys
is the young people so rejoice when G Suite gets rolled out. And one of our bigger customers,
PWC, actually did a poll of their employee base and what
the number one valued thing from particularly their
young employees was their collaboration and
productivity tool. It is so important to them. Okay. So for our next guest, I am really excited to bring out you know, premier
technology company Verizon. They've got 114
million retail customers. We've been working closely with
Verizon for a year now, not simply - and we've been
working with them on workplace productivity. It's been a long project. It's been a challenging project. And that's - you know, I really
wanted to talk about that here. It's gone really well. Great outcome. And here to talk about it,
coming on stage, let me welcome Alin D'Silva who is the Vice
President for Digital Workplace at Verizon. Hi, Alin. Thanks so much for coming. Great to see you. >> ALIN D'SILVA:
Good morning everyone. At Verizon we believe deeply in
our brand statement which says better matters and we believe
better matters not just for our customers but
our employees as well. We are keenly aware of the need
to move faster, collaborate more effectively, have flexibility
in how they work when and where, and it's really a commitment
we've made to a much more modern and progressive
work environment. And we're delighted that G Suite
is going to be a big part of that transformation. >> DIANE GREENE: So are we. We're delighted too, Alin. You know, starting in just a few
weeks, I understand you're going to start moving 150,000
Verizon workers over to the G Suite platform. >> ALIN D'SILVA: Right. >> DIANE GREENE: But we both
know it didn't happen that quickly, that you started down
this path more than a year ago, and I'd just like to know
what were you looking for? >> ALIN D'SILVA: Yeah, so a
little over a year ago we formed a group that was really
focused on raising the bar on productivity through
collaboration and employee engagement and we've done a lot
in that short period of time. We fostered collaboration
by rolling out activity based workplaces. We rolled out unified
communications globally. And we made improvements
to our mobile options. We realized that we were really
missing some key features, things like real-time
collaboration, and other modern features that you see in
these productivity suites. So we did a comprehensive trial
of G Suite and our employees really loved it. They loved the simplicity and
intuitiveness of the product, the ease of use. The real-time collaboration
really blew them away. You know, it
worked well on desktops. They were really surprised by
how - was how well it worked on mobile devices. And you know, the simplicity of
the solution actually helped us secure the platform in a much
more straightforward way. And then I'd say the last
thing is that we realized that actually a number of
our employees at Verizon, specifically at companies like
AOL and Telogis were already on G Suite. So for us, that is how we came
about making the decision. I would say one more thing. Like, you know, we also realized
that we had to skate to where the puck is going because in
terms of talent, we knew that you know, as the years passed,
the talent that comes into the company is going to
expect a product like this. And so we had that in mind
as we made our decision. >> DIANE GREENE: Thank you. You know, Alin, a lot of the
time change can be pretty hard no matter how good
the technology is. Did you have resistance in going
to the cloud and what did you do about that if you did
have that resistance? >> ALIN D'SILVA: Oh, sure. We had a fair amount of
resistance but it really is a matter of addressing the
fundamental perceptions of what it takes to go to the cloud. So what we did is we cracked
open the door, we asked everyone to sort of keep an open mind,
and then over several months, you know, we got really deep
on a number of topics like security, operations, user
experience, regulation and compliance matters, and we also
looked closely at ROI of course. And in going through the trial,
in talking to you, in talking to your partners, in talking to
your customers, we realized that really it was the
right decision for us. The Google team was really
fantastic to work with. Eric Geitner, if you're
out there, kudos to you. But they answered
all of our questions. You know, they got us really
comfortable with the decision that we needed to make. >> DIANE GREENE: Thanks
for that shout out to Eric. Alin, do you have any advice for
anyone in the audience that is looking at making a similar
transformation that you have gone through at Verizon? >> ALIN D'SILVA: Absolutely. So first of all, you've got
to bring along all of the stakeholders that are going to
be key in the decision-making that you're going to make. So bring along your CSO and
legal teams for starters for sure because they need to
kind of go on this journey with everyone. You know, they've got to
understand what it's going to take and get
comfortable with it. The other thing we found is that
we learned a lot from customers, from Google and G Suite, and
from partners, so I would say definitely take
advantage of that. Remember that this is like
partly - I mean, it's a technology transformation but
it's really about change and driving change in the
organization, so you've got to give thought to how you're
going to drive change management around this. I would say, like, you know, if
your mission is to go transform the company, then you really
have to have meaningful change, otherwise, like, where
is the transformation? >> DIANE GREENE: Right. So what has the reception been
like since you announced this move to G Suite? >> ALIN D'SILVA: The reception
has been pretty interesting. So we put out an article on our
intranet site in the middle of January and within a couple of
days, it got like the most reads for the whole month. We had hundreds of comments
and questions that were posted. And you know, people saw it as
a really meaningful step in the right direction. These were folks who had
experienced G Suite at other jobs or had used it in their
personal lives and they were really comfortable and excited
about the idea that we would go forward in this direction. And then there were a number
of employees who you know, naturally were concerned about
how it was going to change the job. They were used to doing
things a certain way. And they posted a lot of
questions online and we went through and we answered
each and every one of them. And that really helped the
broader community understand where we were going and
that was really important. We also identified all of the
early enthusiasts, the ones we, you know, we call them Google
Guide just as Mike said earlier. We're going to rope them into
the early adoption phases of this so that they can get on
board and get ready to support our colleagues when
we bring them overall. So you know, it was
a great validation. It was a great early validation
that we made the right choice and we're really looking forward
to starting the migrations in a few weeks. >> DIANE GREENE: You know, it's
a big step, Alin, and I really admire you know, the commitment
you've brought to bear on this. We've been
working pretty closely. I know I'll be hearing more from
you and I really look forward to the updates and supporting you. >> ALIN D'SILVA: Great. Thanks, Diane, for having me. >> DIANE GREENE: Thank you. Thanks so much. Bye bye. Yeah. Yeah. G Suite, you know, is
built for the cloud, integrated, works seamlessly. It's really proving its value. Going back to Google Cloud,
once someone decides to move to Google Cloud, you know, we
really have an incredible commitment to deliver top
quality customer support. We don't want any of
our customers going down. You know, we use site
reliability engineers, you know, to keep our seven apps that
have over a billion active users available and we realized, as
I said before with the Niantic Pokémon Go launch, that our
customers need the same. We established a new
organization called Customer Reliability Engineering where
those people work with our customers to make sure that your
services are architected and monitored and keep
full availability. Now Google can't scale to do
enough Customer Reliability Engineers for all of our
customers and so what we've done is we've reached out to partners
and I am really excited today to announce two of
our early partners. We have Pivotal Labs, premier
consulting development company leading in agile development. Our first certified CRE partner. They offer CRE
services to our customer. And then another customer,
really a pioneer in the cloud - they understand
the cloud inside out. That is Rackspace and they
are our first managed service provider and we look forward to
doing more and more with them. Now somebody that has already
taken advantage of our CRE, Home Depot. I am really excited to welcome
to the stage Paul Gaffney, Senior Vice President
at Home Depot. Hi. Hi, Paul. Thanks so much. >> PAUL GAFFNEY: Hi, Diane. It's great to see you. >> DIANE GREENE: Yeah. Okay. So you've got an impressive
history, quite an innovative retailer doing extremely well. How do you think
about technology? >> PAUL GAFFNEY: Technology is
a really important thing to us, particularly in this landscape
as consumers' expectations are increasingly being shaped by
the software that they use. We've got over two thousand
stores in North America, 400,000 real live human beings who take
care of our customers and that is our heritage, that excellent
experience that you get when you're talking to someone
wearing an orange apron. And increasingly consumers are
expecting that we do exactly that same thing in software. So we think about
software as an extension of our overall culture. >> DIANE GREENE: It's sort of
your virtual company running isn't it? >> PAUL GAFFNEY: It is. >> DIANE GREENE: And - and what
have been the proof points in using the cloud for you? >> PAUL GAFFNEY: The important
thing about cloud for us as we started to get better as
a software engineering organization - we paid attention
to a lot of different things. One of those things was
building cloud native apps. There were a variety of reasons
why we focused on building cloud native apps but we also said
well we also should probably run them on the cloud if we're going
to build them cloud native. And cloud plays an important
role for us economically. Cloud economics are a
substantial improvement to our on premises economics. But also thinking
differently about resiliency and availability, moving from I
think what is a classical big company, IT orientation which is
make the infrastructure robust, to a more modern view that says
just assume that everything is going to fail. That was a big catalyst for
us to engineer our software differently and partner with you
guys on learning what does it mean to live in a world where
you make your stuff resilient to failure. >> DIANE GREENE: You know, I've
come to love all of these days when you have these spike loads
and you guys recently you know, of course every year
you have Black Friday. I was just wondering; how did
it go for you guys this year? >> PAUL GAFFNEY: I'm glad that
you love these spike loads and - and as shareholders - >> DIANE GREENE: I do bite
my fingernails though. >> PAUL GAFFNEY: So do we. >> DIANE GREENE: Not really. >> PAUL GAFFNEY: As everyone in
here is a consumer, I think you guys all know that there are
times during the year when all of us in retail attempt to
conspire to get you to do more shopping than is normal. And - but those of you who are
in the room who know that that ends up somewhere, that has
to be processed, one of the fantastic things about an
elastic environment is you don't have to attempt to provision in
advance and pay for in advance all of that spike. And if you architect your stuff
correctly, you can scale that out instead of
having to scale it up. And we were really nervous, not
because we weren't prepared. The team that worked on this
does fantastic prep and you guys were a great partner. >> DIANE GREENE: You have
great technologists, yeah. >> PAUL GAFFNEY: Outstanding
prep, but everyone is kind of nervous. Thankfully our 400,000 folks do
such a nice job each year that more folks do
business with us each year. So this last Black Friday was an
all-time high for us and we were running meaningful portions of
our e-commerce infrastructure on Google Cloud for the first
time and it was our smoothest performance ever. >> DIANE GREENE: Just tremendous
to see and it's great to see you using Kubernetes and containers. How about your
work with big data? >> PAUL GAFFNEY: I think as
you know, a very interesting offshoot of our partnership on
cloud was thinking about okay, we're going to be moving some
mainline, line of business workload from on
premises to the cloud. Should we be doing
the same thing with data? And as the audience might
imagine, we've got a lot of data and we replenish those
2,000 stores continuously. And we have started to move
quite a bit of our data analytics environment from
proprietary on premises infrastructure onto big
query and other parts. I think, you know, I'm excited
by the technology, but frankly I'm more excited by the
investments you guys are making in talent and I'm looking
forward to the work we're going to do on exploring some really
interesting problems where we think that the machines can
help us understand some things perhaps that the humans find
difficult to understand. >> DIANE GREENE: Yeah. We're really looking forward to
this ongoing work and I can tell you that Fei-Fei Li and her
cloud machine learning group are really excited about what
they're going to be able to do with Home Depot. Thanks so much,
Paul, for being here. >> PAUL GAFFNEY:
Great pleasure, Diane. Thank you. >> DIANE GREENE: Yeah.
Okay. Bye bye. Thanks. So now we're going to turn
and we're going to talk about security, privacy, data
analytics at scale with a big bank and here to lead that
discussion and tell you about it, I'm delighted to welcome
Tariq Shaukat, Google Cloud's President for all customer
facing operations. >> TARIQ SHAUKAT:
Thank you, Diane. Morning everybody. So you've heard a lot today from
our customers and you've heard a lot today from Diane and from
Sundar about what our mission is here at Google Cloud. But to put it in the terms of
what we're looking to do for customers, we very much view our
mission at Google Cloud and on our customer team as generating
real - a real step change in business value
for our customers. That business value happens in
a number of different ways and increasingly in mission
critical applications. It happens first of all at the
cost savings level of how do you actually move to a more
flexible environment? How do you move to a
more agile environment? How do you move to a
lower cost environment? So that is absolutely one of the
major conversations that we have with our customers. As you heard Paul talk about and
you heard Mike from Disney talk about, we also are increasingly
seeing the cloud being used as a platform to drive real, tangible
revenue growth for companies for them to both create new
applications but also critically importantly, to understand their
customers better, to get into their data better, and to really
understand how to become closer to their customers, how to
become more efficient in their operations, and how to
engage their employees better. So you have the cost savings
piece and you have the revenue generation piece. But the third part that really
has been a thrilling evolution of the market is how many
companies are really coming to Google Cloud now and saying we
need a partnership and we need a partnership to help transform
our business, to really help us think about what is the business
model of the future and how do we take our existing operations
and how do we move that in a both gradual way and in a step
change way into the future? One of the most thrilling
partnerships that we have in this regard is with HSBC. HSBC challenges us in each and
every one of these areas, in the cost savings, in the
revenue growth, and on the transformation levels every day. You may know them as one of the
largest financial institutions in the world. They have over 150 year history,
4,000 offices in 70 countries, and over 2 trillion dollars of
assets on their balance sheet. Now to tell you more about this,
I am truly delighted to welcome to the Google Cloud stage,
Darryl West who is the Global Chief Information Officer at
HSBC, and he is going to share his thinking about the role
of cloud and cloud scale data analytics and machine learning
for the financial services world. Darryl. >> ANNOUNCER: We are living in
a time of rapid digital change, where the only limits are the
scale of our ideas and degree of our dedication. The technology ideas of today
will be the intuitive habits of tomorrow. Never before has innovation
promised so much, touching so many in so many different ways. Our frontier is
financial technology. It's changing the way we work,
make money, and the way we trust each other. To succeed, we need to adopt the
mindset of the silicon hubs of the world and create an
incubator of free thinkers, thought leaders, innovators born
of collaborative spirit, all fueled by a belief that
technology will enhance our customer's financial lives, to
imagine a future beyond what we already know, creating a team
with the power to transform the world's leading
financial enterprise. We are technology. >> DARRYL WEST:
Thank you, Tariq. Good morning everybody. It gives me immense pleasure
to be able to be here with you today, particularly on
International Women's Day. I've been at HSBC now for a
couple of years and I have the honor and privilege of leading
an awesome IT team that is at the center of an amazing
transformation happening in our organization. I'm going to talk about our
commitment to cloud and I'm going to talk about how we're
collaborating and partnering with Google to make HSBC a
simpler, better, and faster organization, an organization
focused on amazing customer experiences through digital
channels, and a company committed to using the latest
technology and data analytics and machine learning to
transform the way we run our business. Now I'm sure many of you in the
audience are travelers and I'm sure many of you have stood in
airport walkways and seen the HSBC advertisements and
they are quite distinctive. And you've
probably asked a question. How are these people? Well, who are we? We're the world's
largest international bank. As Tariq said, we've been around
for one hundred and fifty years. We're present in
seventy countries. We have 37 million customers,
customers ranging from individuals to small businesses,
mid-sized corporates, large global companies, as well
as governments as well. So we have a huge business,
a huge global business. We're the number one bank in
the world for trade finance and cross border financing and we're
a significant player in the global foreign exchange markets. So given our central position
in global finance, as you can imagine, we are a systemically
important financial institution, heavily regulated. In the banking business, clearly
trust and confidence is a central part of our business. We have to make sure that our
customers feel confident and trust in us to be the
custodians of their assets. So you know, information
security, reliability, and resilience in how we deliver our
services are fundamental to our business. Now apart from having the 2.4
trillion dollars of assets on the balance sheet, we also have
at the core of the company, a massive asset in our data and
you know, what's been happening in the last two or three years
is a massive growth in the size of our data assets. So as you can see here, 56
petabytes of data in 2014. That is actually doubled. As of now sitting here, in 2017
over 100 petabytes of data. And what is happening is our
customers are adopting digital channels more aggressively. We're collecting much more data
about how our customers are interacting with us. And obviously embedded in this
data is massive insight and what we need to do as a bank is work
with partners to enable us to understand what is happening
with this data, draw out the insights so we can run a better
business and create some amazing customer experiences. Now our journey in data is very
similar to other companies. This is no different to
many enterprise companies. The history here. We have lots of good old
fashioned databases, all of our core systems are running on
products systems that have been around for twenty,
thirty, even forty years. The systems are robust and
scalable and they do a great job, but they don't have
the database structures that actually allow us to really do
the data analytics, the machine learning that we need to do. So the history of enterprises
like mine is that we typically did extracts into traditional
data warehousing platforms which worked well for many years
but of recent times, they have become expensive to
run and difficult to use. So we took the plunge about
three years ago to really embed ourselves into the
evolving Hadoop ecosystem. And for an organizing like
mine that has this traditional history of what we call legacy
platforms, it's been a tough - it's been a tough road for us. All of this is
being done on prem. We've had to provision,
large infrastructure, physical infrastructure, build out data
centers, and hire talented new people that have understanding
of these new technologies. So we thought about this very
carefully in the last year and we said to ourselves,
we're really - we're a bank. At the core of our business,
we're a bank, but we also have a significant technology company
embedded within the organization and the question that I was
asking the management team is you know, do we really want to
compete with the cloud providers and people like Google? Are we really going to try and
do what they do as well as they do it? I think our conclusion was that
it was better for our business if we adopt a
cloud first strategy. So we started working with
Google about six months ago now. So it's still early
in our relationship. But we had been working
with them on some of the most important and critical business
problems that we have to solve in our business. To give you an idea of what
these types of problems are - so the initial use cases here
are typically characterized by business problems that have very
large data sets and require very intense computing
capability in short bursts. Now the first one on this
list is all about anti-money laundering so as one of our
obligations of being a bank, we worked with the governments and
the crime agencies to identify nefarious activity
and money laundering and criminal activity. Now we have a set of
applications that monitor a huge time series of data, so a
massive dataset, billions of transactions for
all of our customers. Running analytics over this
huge data set with great compute capability to identify patterns
in the data and to bring out what looks like
nefarious activity within our customer base. Those patterns that we identify
are then escalated into the agency, so we work with them
to track down the bad guys. So this is an application that
clearly requires massive data sets, great computing
capability, but also a machine learning capability to be able
to identify the patterns in this huge data set. The other applications that
are mentioned here in terms of finance and risk, we have
billions of transactions. Those transactions need to be
aggregated so we can manage our finances and our
risk at a global level. Evaluation services and our
trading business, as I said, were a major player in
global financial markets. We need to be able to run
complex multicolor simulations on a regular basis to be able
to better understand our trading positions and our risk. So this requires a
significant compute capability. So we were faced with the
question of do we build that new data center and put thousands of
servers and thousands of cores out there to do this activity,
or should we actually work with somebody who does it
as their core business? Our solution was rather than
build that out ourselves, we should be working with you know,
world-class leaders in this space, and hence
our work with Google. Now all of this doesn't happen
by accident and there are some critical enablers that have to
all be in place to make this work for organizations like us. The first couple of boxes here
talk about risk and compliance and information security. You know, we are a
heavily regulated business. We have to provide a very
resilient, very safe, and very performant services to our
customers and also very agile because the requirements of
our customers are changing tremendously quickly. We also have to respect the data
protection and data residency rules that we have in
our seventy countries. There are lots of rules about
what data can sit where and what data can be shared. So we have to
respect all of that. Working closely with
our partner at Google on this - on this construct. Information security as you
heard this morning, it's a massive issue for everyone,
particularly a bank. We have to keep our data safe
from cyber criminals and we are working very well with Google on
solutions in the security space. Probably the biggest impact
though or the biggest enabler for this journey is the other
two boxes in the middle here. You have to drive a cultural
shift in your business to make this really work. You have to adopt an
agile methodology. You have to adopt
a dev ops mindset. And you need to be able to
recruit and retain talented people that understand how to
use these new technologies and working this way. So just a point of
note; we are recruiting. Okay. And we're a great
company, just like Google. So the other issue here on the
second box is of integration. I mentioned the legacy
platforms we have. All of our core data is sitting
in these systems of record. We have to make sure that we can
integrate all of this technology complexity to work with the
Google platform, the Google Cloud platform. So this integration layer
is very, very important. You've got to get
that culture in place. You have to get the people that
understand how to do it and you have to build the integration
layer to be able to make it all work in that rapid cycle. And lastly of course,
data preparation. You know, we have, like most big
enterprises, a data sprawl as I said and a massive growth of
data, but also you need to invest your time and effort
in your data architecture and making sure you understand where
your crown jewels are, where the copies are, and get a really
good understanding of the preparation of getting the data
ready to be able to interface effectively into
the cloud partner. And obviously agreements
of legal and commercial are obviously very
important as well. So in summary, these enablers
are fundamental to us achieving our goals at HSBC and we are
working in very collaborative and partnership mode with Google
to get these enablers in place to run a better business. >> TARIQ SHAUKAT:
Thank you, Darryl. That is a fascinating journey
that you guys are on there at HSBC. Just a couple of
questions for you. You mentioned that you did a
pretty thorough investigation of the technologies, in particular
as you were looking at us around data flow and big query. What aspects convinced you
to move forward with GCP? >> DARRYL WEST: Yeah, we've -
we spent a good amount of time working with all of the big
cloud providers and evaluating their products. I think if you look at the
scarce talent we have on data analysts and data scientists,
they need tools that allow them to do their work in a
very productive way. So we found as we worked with
your team and your subject matter experts, working with our
data scientists, that your suite of products have been very
performance - very easy to use. You know, there is always
a challenge of getting data scientists to shift from their
favorite platform to a new one and I think the way that your
products are actually being configured and presented to
our scientists, they've really enjoyed that experience. So you know, it's
early in our relationship. We're still in pilot mode and we
hope to get these use cases to production in the coming weeks
and months, but I think our experience so far
has been very positive. >> TARIQ SHAUKAT: Terrific. Just looking forward the next
three to five years, how do you see the cloud really playing out
and all of these technologies impacting your business? >> DARRYL WEST: Yeah, look, it's
very difficult to be able to predict the future,
even three years ahead. It's a long way away, right? So the world is going to change. I think what is definitely going
to be consistent is that our data is going to
continue to grow. Massive data sets
need to be managed. I know also we're going
to be pushed to go faster. I mean, our customer base is
dragging us into a new world and consuming services on new
digital channels which will require us to be
agile and fast to market. I know we're going to have the
need to use data analytics and particularly machine learning
to run a better business. So it's going to get bigger. It's going to get more complex. And we're going to need to
partner with people like Google and others to be able to use the
best technology in the world to solve these
problems for our business. >> TARIQ SHAUKAT: Great. Well finally, I think everyone
in the audience I'm sure could use a little advice. You guys are leaning very
forward into the cloud and cloud technology. What advice would you give
everyone here for how to really start their journey? >> DARRYL WEST: Yeah. Well I think the first thing you
need to do is you have to make the jump in thinking to make
yourself a cloud first company. It requires you know, getting
everybody on board in the technology team,
but also the business. You've got to get the business
to think about things in a very different way and I mentioned
the criticality of new ways of working with agile methods,
dev-ops methods as really important. So that first thing is make
that jump and be bold and have courage to go and do
the innovative things. The second thing is pick a
partner that has the same type of culture as you. I think you need to. If you're going to be
innovative, you're going to be bold and take on big challenges,
you need a partner that is going to think the same way and work
in the same ways of working. And just I think basically the
conclusion I've come to is have a good look out there. Meet, you know, the people in
the market because they are all very different. And certainly from my
perspective, the chemistry between the people and the teams
is actually the number one thing because what we're
doing is complex. What we're doing is difficult in
many cases and the chemistry and the culture of the
organization is very important. >> TARIQ SHAUKAT: And
it's a multi-year journey. >> DARRYL WEST: And it's
a multi-year journey. So you have to like who
you're hanging out with for the years ahead. >> TARIQ SHAUKAT: Exactly. Well again, on behalf of all
of Google Cloud, we're really thrilled that you're here
today so thank you very much. >> DARRYL WEST: Thank you. >> TARIQ SHAUKAT: And I'd like
to welcome Diane Greene back onto the stage. >> DIANE GREENE: Thank
you, Tariq and Darryl. That was fantastic. And now we're going to have a
look at how an iconic e-commerce company, EBay, is
thinking about the cloud. I think we have a video first
and then I want to welcome on stage RJ Pittman who is
head of products, CTO. >> ANNOUNCER 2: We transfer
about thirty gigs a second, search through one billion
active listings in our quest to find your perfect. Through cutting edge technology
and the performance and scale of cloud, we strive to meet our
customers where they are and transform the future of commerce
to help find your version of perfect. The future of
commerce from EBay. >> DIANE GREENE: Cool.
RJ. Thank you for coming. Great to have you here. >> RJ PITTMAN: Great to see you. >> DIANE GREENE: So, RJ, you've
had a phenomenal amount of experience working in
online environments. Why don't you start talking
about how your view of the cloud has changed? >> RJ PITTMAN: Well the biggest
thing is it's moving and it's moving fast. In the last four years, it's,
from my perspective, gone from a very powerful and compelling
proposition for outsourcing your IT infrastructure and shifting
some things so that you can focus on your core business,
to now we see it as a strategic growth engine and a
strategic growth engine that accelerates innovation. This is a very different
position that the cloud has taken and it is playing a very
important role for us in that matter at EBay. >> DIANE GREENE: And what do
you think that means for EBay? >> RJ PITTMAN: Well you know,
in the videos you can see. I mean, we've been
around for a long time. The business is
continuing to grow. We have over one billion live
listings at any given moment in two hundred countries
around the world. >> DIANE GREENE: Yeah. >> RJ PITTMAN: One hundred and
sixty-seven million shoppers. And so you can imagine that if
we start moving the capabilities of the cloud to EBay, we're not
just changing our infrastructure and changing our business; we're
changing the experience for all of those shoppers and all of
our buyers and sellers around the world. And that is the really important
connective tissue here. This is not just a
technology exercise for us. It is absolutely a
customer-centric effort. >> DIANE GREENE: Yep. Absolutely and you know, with
that kind of phenomenal growth, I can see that you really are
paying attention to technology to help you through it. Yeah? >> RJ PITTMAN: Yeah, it's -
you know, for us the e-commerce space is competitive and it's
moving fast and the demands of customers and shoppers today is
changing so quickly and as new generations of buyers come into
the world, first millennials and now gen-zers, they have very
different shopping habits, very different expectations, and they
are much higher expectations. So we EBay, at our scale,
if we're, you know, intent on setting ourselves up to meet the
demands and exceed the demands of this next generation of
online shoppers, we have to move quickly and the cloud has become
a very strategic partner for us in that way. >> DIANE GREENE: So RJ, before
we get into our cold demo, can you just tell me a little bit
about what you're doing on Google Cloud? >> RJ PITTMAN: Sure. So this was an undertaking for
us that we've been thinking about for a few years, but we
really got serious about it a year ago and we knew that there
was a lot of untapped potential inside the EBay marketplace
and we wanted to get at it. And as a marketplace, let's
also remember that first and foremost, we're a tech company
and we utilize technology to differentiate, to compete to
innovate, and to create these amazing customer experiences for
all of our customers around the world. And so for us, we went out on
a mission and said if we could bring the EBay marketplace to a
modern stack environment, if we could bring our customers to a
place where we can be innovating almost every day and bringing
great new features and capabilities to them, you know,
that would give us a tremendous advantage and a ton of momentum. So literally in the space of
about five months, we took a great team of engineers and
set off to do exactly that. And in five months we went
from nothing to the entire one billion live listings
of EBay running in GCP. It was not a test. It was not a prototype. It quickly moved to a production
grade environment for us that we ended up launching and bringing
in to market six months ahead of our plan and mind you that, in
total start to finish, was about seven months. So it can be done. >> DIANE GREENE: Yes. >> RJ PITTMAN: Do
not fear the cloud. It can be done. And look, in our business
this is super important. You can't - you have to be
careful when you're serving live customers in a live
marketplace like that. And now the way we've approached
it, we were smart about this and we didn't just flip the switch
and you know, move the entire 85 billion dollar marketplace
business at the throw of switch. We're actually running a
parallel swim lane and that gives us lots of
latitude to play with. They're both live systems and
this is very difficult to do in - in most environments because
we're transacting as you saw, thousands of orders a second. And we now have two hot,
hot systems running. Our core system on prem and then
we have the new Google platform both serving customers and
we've had to build a system that ensured that we didn't create
two separate marketplaces, meaning no two buyers could buy
the same one item that is for sale because that would create
obviously a lot of problems for us. And it was actually one of the
most important decision factors for us in really leaning in
on the Google platform is the services and the capability that
Google offers that nobody else did, that allowed us to build
a true replication - replicated system that runs in real-time
side by side to help companies of our size and scale and
complexity make that transition much more
seamlessly to the cloud. Now mind you, the other cool
thing about it is it didn't just allow us to go to a cloud
instance here in North America, but using the global footprint
of GCP, we took the show on the road and we were able to test
our theory and look at lighting up the Google Cloud in multiple
countries around the world so that we could be serving where
our customers are and providing a great high speed performance
as if that data center was sitting in their backyard. We've not been able to offer
that kind of experience and that kind of performance in the
twenty-two years that we've been around. So this is really breakthrough
for us and it's our customers that are the ones
that really stand to gain. >> DIANE GREENE: That is just
a great story about what you've done and I think also in another
area, you're kind of leveraging AI and - >> RJ PITTMAN: Yeah. >> DIANE GREENE:
For your catalog. And what about that? >> RJ PITTMAN: So this is - this
is a really cool part of the story where you have a billion
live listings and we have twenty-two years of
historical transaction data. We have more economic data
on the supply and demand information on virtually any
product that has been made and sold in the last twenty years. We know what it sold for, what
the demand was, what countries, you know, are most interested in
what products and when, and you can imagine if we could then
start to encapsulate and capture all of that knowledge and bring
it forward for our shoppers. You could build arguably
one of the smartest shopping experiences and smartest
catalogs on the planet. And the challenge there too has
been getting to a place where we had truly the computing capacity
and performance to get our arms around literally billions and
billions of records that go back in time but also the new ones
that are created every day that there are transactions on EBay. And going to GCP, we have the
scale and the performance that allowed us to lay down some
really great AI technology, going well beyond machine
learning and going very seriously into deep learning,
natural language understanding, and computer vision to create
something that we called EBay ShopBot. This thing is pretty great and
it is the largest marketplace in social where we took our billion
item catalog and merged it with a billion people in the social
graph and brought them a highly personalized shopping assistant
that gets to know you, that understands natural language,
and can start to make really, really great recommendations
across our entire inventory of a billion items to give you - help
you find the best item for you at the right time, at the right
place, at the right price. And so when technology can start
doing the heavy lifting and provide you with almost a
concierge-like personalized service, I think that's when
we've really, really made great strides, and that's something
that we brought together with the innovation at EBay and the
great work going on in Google. >> DIANE GREENE: I have to say
that sounds like a perfect match for Google Home and we actually
have a Google Home sitting over there. Could you show us all
what you're talking about? >> RJ PITTMAN: In the spirit
of going to where our customers are, we've taken it to Google
Home because we want to get inside the home. Let's do it. >> DIANE GREENE: Now you
know, we're doing this live. There is like ten thousand
people on the network. We really hope this works. >> RJ PITTMAN: Hey Google. Let me talk to EBay. >> GOOGLE HOME:
Sure. Here's EBay. Hi. I'm EBay. I'm the world's price guide. You can ask me what
something is worth. >> RJ PITTMAN: What
is my camera worth? >> GOOGLE HOME:
Is it new or used? >> RJ PITTMAN: New. >> GOOGLE HOME: What is
the brand of your camera? >> RJ PITTMAN: Canon. >> GOOGLE HOME: Tell me the
model of your new Canon camera? >> RJ PITTMAN: Five D. >> GOOGLE HOME: What else do
you want to tell me about your Canon camera? >> RJ PITTMAN: Nothing. >> GOOGLE HOME: Your camera
is worth 2,548 dollars. >> RJ PITTMAN: Not bad. So this is maybe an unexpected
use case, but one that is really interesting and close to the
hearts of EBay because we support not only our shoppers
but also our tens of millions of sellers all around the world. And what this allows us to do is
start to bring more sellers into the marketplace. You may have lots of really cool
things sitting around your house that you didn't know had
really great market value. Now you'll be able to ask Google
and ask EBay through the Google Home partnership to get you
started and bring you a little bit closer to EBay. And there is so much more to
come with this technology. >> DIANE GREENE: Absolutely. >> RJ PITTMAN: And
let me tell you this. It was five months for us to
go from where we were to in the Google Cloud and it took us five
days to get into Google Home. So you know, with that
kind of power there is great, great potential. >> DIANE GREENE: RJ,
thank you so much. >> RJ PITTMAN: Thank you. Thank you. >> DIANE GREENE:
That was fantastic. >> RJ PITTMAN: It was
really fun being here. Thanks everyone. >> DIANE GREENE: You know, I
just love the fact that here is a chat bot running on Google
Cloud, talking to Google Home. We have one Google
partnering with EBay. And now you're
all in for a treat. We're going to bring
Fei-Fei Li on stage. I first met Fei-Fei when she
came to Stanford back in 2012. She is a neighbor of mine. And for those of you who
don't know, Fei-Fei was really instrumental in the
explosion of machine learning. She built image net which is
the largest database of labeled images and what that facilitated
was all of the research - all of the academic researchers,
industry researchers now had something to benchmark against
and we started seeing this explosive improvement and
adoption of machine learning. So let me just welcome
to the stage Fei-Fei Li. She is the - she is in charge of
Google's cloud machine learning. She is head of the AI Lab at
Stanford and here now to lead us in our AI efforts at Google. Please welcome. >> FEI-FEI LI: Good
morning everyone. My name is Fei-Fei Li. I'm the Chief Scientist
of Google Cloud AI and Machine Learning. So in Google's code word, I
am still a New-gler, and it is quite an honor and privilege to
be on this stage to share with you some of my thoughts
about AI, machine learning, and Google Cloud. So the world is
changing incredibly fast. Some say that we're living
in the fourth industrial revolution, and much of it is
propelled by the phenomenal force of computing. As an AI technologist for nearly
twenty years working on machine learning, computer vision, I
have witnessed my field growing from a lofty but academic
pursuit to the biggest driver of this change. But change happens at many
scales and it takes imagination to see them all. Let's take a familiar
example, the self-driving car. It is easy to
understand the appeal. With the help of sensor and
algorithms, a self-driving car reduces accident risks and gives
us more time to work, socialize, and just relax while commuting. This is great for
a single driver. But what happens when
thousands of people have self-driving cars? Suddenly through the
coordination of vehicles, traffic congestion
is reduced and parking is dramatically simplified. What about millions
of people having them? Entire cities will be reshaped
by - to reflect the fundamental shift in the use of
its infrastructure. So the difference between
each scale is participation. As a technology reaches more
people, its impact becomes more profound. This is why the next step for
AI must be democratization, lowering the barriers to entry
and making it available to the largest possible
community of developers, users, and enterprises. Speaking of democratization and
reaching many people; Google's Cloud Platform already delivers
our customers' applications to over a billion users every day. That is a lot of participation. Now if you can only imagine
combining the massive reach of this platform with the power
of AI, making it available to everyone. Then we stand to witness a
greater improvement in quality of life that any other time
in history from finance to education, from manufacturing
to healthcare, from retail to agriculture, you name it. This is why delivering AI and
machine learning through Google Cloud excites me. It means finally sharing the
technology and insights I have been involved in for years as
an AI researcher at Stanford. That is also where, by the way,
I began a collaboration and partnership in AI with Dr. Jia
Li who was one of my first PhD students many years ago. I am very excited that she has
joined Google with me as the Head of R&D in AI and
Machine Learning at Cloud. And speaking of International
Women's Day today, this is another badass woman
in STEM, CS, and AI. There is no shortage of examples
of AI Solving real world problems such as the demo we
just saw that EBay - talking to the EBay ShopBot
through Google Home. As impressive as these
achievements are, they are just the beginning of the
transformation of the entire enterprise. More and more problems are being
addressed by AI and the tools we use to build AI solutions
are becoming more and more sophisticated in their
functions, but also easier to use. This will change our world
dramatically and it is happening at a faster pace
than most people think. Let's take a look
at a few examples. AI has been influencing
retail as long as it existed. For instance, machine learning
algorithms are already helping Google's AdSense and Shopping to
deliver relevant information to our customers. But so much more is just waiting
to be done such as supply chain optimization for routes and
inventory, or predicting changes in demand over time. Drones navigation and
self-driving vehicles for delivery of items
customers ordered. Intelligent analysis for
loss prevention and safety. Or understanding customer
movements and perception in stores to optimize shelf space,
displays, and visibility. Another example from media and
culture already being influenced by AI. Do you have a teenager at home? Wonder why - what is the
technology that is mesmerizing them with the cat ear and
rainbow filters of the Snapchat app? That is a clever
computer vision technology. Machine learning already
delivers Google Photo automatic image tagging and
YouTube's recommendation list. But media experiences will
soon make much more use of it. AR and VR will rely on computer
vision for motion tracking, environment
detection, and games. More and more news content will
be automatically generated, allowing journalists to focus
on bigger and deeper stories. And AI will play a growing role
in helping us to create and stylize our own content, such
as videos, music, and artwork. In the financial service world,
we are already seeing machine learning to help fairly and
intelligently to predict credit card risk for new applicants or
even anticipate delinquencies among existing customers. And many similar advances
are in the work as we speak. Insurance claim will be assessed
by machine learning agents. Banking will become even more
virtual as conversation bots take over call centers or
even in person bankers when managing finances. And as HSBC said earlier, our
own perception will be augmented with intelligent agents to flag
criminal activities such as money laundering or fraud. Last but not the least example
here, healthcare is among the most profound applications
of AI for truly improving peoples' lives. We have already seen some
incredible AI achievements in recent years. A few months ago, my colleagues
at Google Brain have shown that using a deep learning algorithm,
a computer can detect signs of diabetic retinopathy, a disease
that could potentially blind more than 400 million people. Now imagine that kind of
insight spanning the entire healthcare industry. So many forms of visual
diagnosis may soon be automated to help doctors and reduce
overhead and errors, and extending treatment to
the underserved population. Machines can also help to handle
clerical tasks such as scribing doctor visits, managing chronic
diseases, leading to more reliable and faster services. And this will accumulate into in
full scale smart hospitals and homes with intelligent sensors
to track hospital activity, keep patients safe, ensure adherence
to hygiene practices, and augment surgical protocols. I hope you're excited just as I
am about the opportunities that AI and machine
learning can deliver. But this remains a
field of high barriers. It requires rare expertise and
resources few companies can afford on their own. That's why cloud is the
ideal platform for AI. That is also why we're making
huge investments in cloud AI that will emerge over the next
year in the form of powerful, easy to use tools that will give
every cloud customer an app into this field. In other words, Google
Cloud is democratizing AI. This entails four broad steps,
democratizing computing, democratizing algorithms,
democratizing data, and democratizing
talent and expertise. Let's talk about each of -
what each of these means. First and foremost, AI
requires enormous computing. Today, a deep learning algorithm
that can easily boast tens of millions of parameters and
billions of connections. Training and using
such models requires computational resources. Of course that is exactly
what the cloud was designed to deliver. Last year we introduced the
beta version of Cloud ML Engine. Today I am here to announce
its general availability. Cloud ML Engine is a platform
that can harness all of the power - computing power and
deliver it to you transparently. Simply put, you develop the
machine learning models however you like using familiar tools
like the TensorFlow Library in your own environment. ML Engine allows you to focus on
the creativity of your solution and leave the
infrastructure to us. Then when it's time to train
those models, upload them to the cloud where ML Engine can
do it much faster and at a much larger scale. Finally, deploy the result
anywhere from your own premise to a mobile device where it can
put its training to use to solve real-world problems. But even with the compute power
in the world, AI remains among the most complex field in
computer science, and that is still a serious obstacle for
many enterprise and customers. For developers not quite ready
to build their own models, the easiest way to put AI to use
today is through one of the APIs Google has provided to deliver
fully trained machine learning models to tackle
common problems. These APIs are like a switch
that immediately activate an intelligent component in any
application, allowing it to understand speech and photos
or translate text or parse natural language. But there is a lot more to
Google's depth and breadth of AI technology. At Google, we have numerous
research teams housing an enormous amount of ongoing AI
research, spanning many areas of AI and machine learning. Our researchers are some of
the most prolific authors of scientific papers at top AI
journals and conferences and our teams are frequent winners of
best papers and AI competitions. And the results of this work are
quickly turned into products and services that we want to
deliver to our customers. So I am especially eager to
announce some of the latest products of that effort. The Vision API has been under
steady development and features some significant
new capabilities. First is an expansion of the
API's metadata to recognize millions of entities from
Google's knowledge graph from images that are
available on the web. We are now using the same
metadata that powers the entire Google Image Search. Second, an enhanced optical
character recognition capable of retrieving text from images of
text heavy documents such as legal contracts and
other complex paperwork. But the world of pixels
go beyond just pictures. In fact, videos are among
the most prevalent form of internet data. YouTube alone sees hundreds
of hours of videos uploaded every minute. Understanding the rich content
of videos has been a tremendous technology
challenge for many years. In fact, many of us computer
vision researchers, we have often considered videos the dark
matter of the digital universe. Today I am very excited to
announce an entirely new API powered by machine intelligence,
the video intelligence API. I would like to introduce my
colleague Sara Robinson to demonstrate this
API in more detail. Sarah. >> SARA ROBINSON:
Thank you, Fei-Fei. Thank you. So the best way to experience
the video API is through a live demo. And let's start by taking this
video of a Super Bowl commercial for Google Home. And I'm going to play
the first few seconds. We can see that it starts with a
mountain landscape and we see a house, a city street, then it
goes to a dog and a garage. So lots of scene changes
happening in this video and if we were to manually categorize
what is happening throughout it, we'd need to watch the entire
thing and write down what was happening in every scene. Luckily the video API takes
care of this for us with just a single REST API request. So it tells us two things. One, at a high level it tells
us what is this video about and then it also tells us at a more
granular level what labels it finds in the
video in each scene. So if we scroll down here, we
can see it identifies a dog and it can tell us exactly where
in the video that dog appears. It also identifies at the end of
the video, there is a birthday cake and if we scroll down a bit
more, we can see that not only does it know it's a dog, it
knows what type of dog it is, that it's a dachshund. And if we scroll down through
the rest of the labels, we can see that it also successfully
identifies that mountain pass scene from the beginning. So this is what the API can do
with one video, but you likely have more than one video
that you want to analyze. So let's take a look at how
a company might use the video intelligence API. A media publisher could have
hundreds of petabytes of video data sitting in storage buckets
and one common thing they might want to do is create a highlight
reel focused on a specific type of content or search their large
library for a specific entity. So let's see how we would use
the video intelligence API to search a large library of videos
given all of this metadata that we get back from it. So we've got a lot of videos
here and let's say this media publisher has hours of sports
video but they only want to find the content
relevant to baseball. So let's go ahead and
search our library here for baseball videos. And we can see that not only
does it show us which videos have baseball, it tells us
exactly when in those videos baseball appears. My favorite example is this one. We have this video which only
has a tiny bit about baseball but it's able to identify that
clip for us, whereas if we were to manually do this, we'd have
to watch the whole video looking for that specific scene. So if you click on this scene,
we can see that this is from last year's Year in Search
video when the Cubs won the World Series. So let's do one more search. I live on the East Coast where
it's pretty cold right now. I have heard there's been a ton
of rain in SF this past year. I think we can all agree it'd
be pretty nice to be on a beach right now. And while machine learning can't
take us there, it can do the next best thing and find all
of the beach clips in our video library for us. So let's search for our beach
videos and then we can click to all of our beach
clips in the videos below. So as you saw through this demo,
the video intelligence API makes it easy to quickly and easily
understand a large library of video content, something that
was almost impossible just a few months ago. Tasks that used to take hours
now take seconds with the video intelligence API and I am
excited to make it available to all of you today. Thank you. I'm going to hand
it back to Fei-Fei. >> FEI-FEI LI: Thank you. Thank you, Sara. So I am so excited to see this
as a computer vision researcher. I have seen my field working on
video understanding for decades and now finally we are beginning
to shine light onto the dark matter of the digital universe
and provide values to our customers who can use it to
harness the enormous amount of information
embedded in these videos. Now let's go on. Data is the next piece of
the democratization puzzle. Just as we learned through a
lifetime of exposure to the world to gain our human
intelligence, AI requires a huge amount of data to
develop its own insight. But these data sets are
among the steepest barriers to overcome. I have a lot of personal
experience of this, having led the development of the image
net data set which provides over fifteen million label images to
the machine vision community. Many of you are now
familiar with the history after image net. In 2012, it became one of the
most important enablers of the deep learning revolution. To this day, it is still one of
the most used training data sets and benchmarks of deep
learning algorithms. While the results of image net
have been incredible, the long, difficult journey of putting
it together was a tremendous testament of how great the
challenge still remains. What we need is a more scalable
and effective way to democratize data to more data scientists,
machine learning developers, domain experts, and
eventually to our business. That is why I am thrilled to
make the next announcement, Google Cloud's
acquisition of Kaggle. For years, led by co-founders
Anthony Goldbloom and Ben Hamner, the Kaggle team, has
been building an unprecedented community of over 850,000 data
scientists, hosting contests, and making new data
sets publicly available. By merging with Google Cloud
Platform, we're giving this community direct access to the
most advanced machine learning environment as well as
providing a direct path to market their models. Working together with Kaggle,
we're empowering the largest concentration of machine
learning talent in the world. In fact, Kaggle is already
partnering with Google Cloud and research to host the largest
video understanding competition called the YouTube 8 million
video understanding challenge. So speaking of talent and
expertise, we're being - we've also been extremely committed to
help our partners and customers to develop more machine learning
and AI expertise at the levels they need. At Google, we've always
made significant investment in research. Every year, Google gives large
grants to over 250 academic research projects worldwide,
supports dozens of PhD students, and hosts thousands of interns. Moreover, the Google Brain
Residency Program recognized that expertise in AI will be an
increasingly important resource in years ahead and is taking
steps to find, educate, and empower the future
leaders of this field. At Google Cloud, in parallel
with all of these efforts, we're equally committed to using
our expertise to deliver real results right now
to our customers. The advanced solution lab allows
customers with more ambitious goals to partner directly
with Google to solve complex AI problems. Let's take the insurance company
USAA as a recent example. Many of their engineers were
well versed in data size and some even had a
background in machine learning. But they needed help to build
a true foundation of expertise. To do this, a team of USAA
developers came to Google's Advanced Solutions Lab where
they learned directly from our own machine learning
engineers and experts. That team now is hard at work
putting their new skills to use with additional teams being
trained in the same way. So I think the most meaningful
technologies are the ones that transform a precious
resource into something that can benefit everyone. The printing press helped
literacy expand beyond the privileged, made books so
affordable that they could fill the shelves in the homes and
libraries all over the world. The electric grid delivered
power to the entire communities, turning heat and light
from luxuries to staples of everyday life. The mass production of the
industrial revolution and the artisanal objects that were once
prohibitively expensive could now enrich the lives of
hundreds and millions of people. And of course the internet has
made everything from newspaper to university courses so easy
to share that they can reach a worldwide audience
overnight, often for free. What these examples have in
common is the transformation from exclusivity to ubiquity. I believe AI can deliver this
transformation at a scale we've never seen and imagined before,
to help spread the luxuries of the privileged few to the
rest of us at a global scale. This is why I am inviting
everyone in this audience to be part of this. We at Google Cloud are making
the tools available but it's up to you to put them to use. And speaking of AI, it took a
few years to convince our next speaker of the power -
the sweeping power of AI. So if you're still on this
journey, it is really not a shame because you're
in very good company. It is my honor and privilege
to invite Mr. Eric Schmidt. >> ERIC SCHMIDT: Thank
you very much, Fei-Fei. >> FEI-FEI LI: Hi. Good luck. >> ERIC SCHMIDT: Thank you. Fei-Fei said leave the
infrastructure to us and I don't think anyone could
have said it better. Last year when I was here, I
said we will meet you where you are. What I thought about this year
is just get to the cloud now. Just go there now. There is no time
to waste anymore. There are lots of reasons and
I think we've talked about them this morning, but imagine a
model that goes something like this. You're sitting there and you
are building something on a container on your laptop. You get it working and then you
just release it to the cloud and it scales infinitely. That is how easy this is now. That is how easy it is
for you to do this as an individual programmer. Now does this matter? Well it mattered
in the last year. Let's think about
Pokémon Go, right? So here is a fantastic product
which had fifty times more demand in its first two hours
than ever planned in their most optimistic forecasts, and I'm
pleased to say that our system expanded, right, that the
container architecture and the use of Kubernetes actually
could handle the load. And I'm quite convinced that
there was no other way to handle such a global phenomenon. And you sit there and you go
well, I'm not trying to be Pokémon Go. Well if you could,
you'd be pretty happy. Think about the
sort of global success. So you might as well plan for
global success and infinite demand, and even if you don't
hit it, your architecture will be alright and your
costs will be lower. I'll give you another example. Start - well if it sounds
like open source, right, using Kubernetes, Apache Beam,
Spinnaker, using GRPC, Redis, HBase, you're
going to have a blast. It's going to be a
great week, right? I'll give you another
example consistent with this. We just had the strongest IPO in
a long time in tech and snap and Snapchat. As part of that, when you read
the narrative, there are two things that strike you. First was their incredibly fast
software development in their app both on Android and iPhone
and in their community with huge success, new models, new models
of interaction which you all know and many of you use. But the other confusion that is
interesting in their process was nobody could figure out how they
could do this with so little capital and the answer
is because they used our infrastructure, right? In a deal that is a
hugely successful deal for both corporations. So you sit there and you go
well why would they do that? I mean after all, had they spent
two billion dollars in data centers, they would have two
billion dollars' worth of data centers. Yeah, but then they'd be putting
three billion and four billion and five billion and six billion
in whereas this way they can ride the scale and the
investment that we have done. So are you not planning
to be like Snapchat? Well I think everyone here who
represents a corporation would be happy to be as
successful as they are. It makes sense, right? But the interesting insight is
in the last year we have done two things since this conference
than I am particularly proud of. One Fei-Fei talked about which
is the sort of making machine learning real at a commercial
and enterprise grade scale. You saw the demos. You understand the message. I'll talk about
that in a minute. But the other thing that we did
is we took the Google network which is vastly better than
anything else you've ever seen, all that dark fiber, all
that special fiber, all the interconnect, and we made it
directly accessible to the programs that you're using. So using Snap as an example, not
only can they use our platform infrastructure but they get
access to their global customers through our network by
virtue of the way this works. That is a very big deal. So you say well, I'm only trying
to get regional customers. Dream big. Think about having customers
in pretty much every country. You get the idea. So what I want you to do is I
want you to sort of understand that we can do both great cloud,
that is great technology in the cloud, but we can also now work
with you as a customer and a partner to build a business that
actually scales globally and which you make a lot of money
which is ultimately what everybody cares about. So I want you to imagine that
- and the good news is, that Fei-Fei talked about and the
others have talked about it as well - we have this incredible
leadership platform for the things that I think
matter for the future. But the truth is that most
of our customers are still struggling with the
enormous challenge of their existing infrastructure. Most customers say that is
a great speech but I'm still trying to deal with my non-X86
apps, my AS400 mainframe, all of the kind of stuff that they have
and many of you work in such companies or run them. So what - how do you do that? So let's imagine a sales call
where I am your salesperson visiting you and we say
what are we going to do? Well I'll tell you the first
thing to do is take your X86 binaries that you have in some
data center inside your company and move them to the Google VM. And you go huh. Well it turns out it's cheaper
and if it's cheaper, that frees up some money to
do the next step. So the first thing
is the easiest one. Literally take the existing PCs
and existing PC software that is X86 software specifically, and
put it under the Google VM. What do you do with
the existing systems? You leave them running. Makes perfect sense. That's phase one. What do you do next? What you do is you
take that system. Now you've got it working. Half in the cloud, half not. And you say I am now carefully
going to take the data out of all of these systems and put
them in a modern scalable managed database,
either SQL or non-SQL. How do I do that? Well we work
with you to do that. And then of course overtime,
you build and move that data to modern end to end services. I'll bet my - the rest of my
professional career that the future of your business is
big data and machine learning applied to the business
opportunities, customer challenges, and
things before you. It's true whether it's the video
API and intelligence service we just announced or more
traditional classification or the new deep learning approaches
that are being pioneered at Google and elsewhere. So what happens is at that
point, you say okay, I managed to get the data over here and
now I don't know what to do. Well Google has a lot of
resources that we can throw at this. We have customer
engineers ready to help you. We've got an office of the CTO
to help you with technology. We've got solutions architects
who actually do a whole business solution for you. We have an advanced solutions
lab that actually showcases some of these things. We have professional services
that you can work with in the ways that you would imagine. We have people who can - are
called strategic customer engineers who help
you build your cloud. And we finally have customer
reliability engineering to keep the thing running, right? So not only is it used to be
we would say well we have this incredible technology derived
from the way Google operates but we didn't have a full
service product offering. We didn't have
all of the service. We didn't have all of the
people who could help you. And you're sitting there going
well, do I really believe you? Well we have the references now. They are among you and you've
heard some of them today. And what's interesting is that
not only do you have to do that but you're going to need
additional help, especially with the legacy systems. So we have system
integrator partners. Examples that I wrote
down, Pithion, Agosto, Guvnats, Sata. Global systems integrators
like Accenture and PWC. So okay, that sounds
like a pretty good pitch. The last time anybody looked
at this, it takes years to make this transition. So that was another problem
that we faced a year ago and we figured out how to solve that. We're working now to get these
conversions to occur in one month, two months, three
months, four months. Why? Because time is everything and
you - while you're doing this conversion, to some degree some
of your customer experience software, some of your
value added is sort of being held captive. So - another example, we
have Lush which is a cosmetics retailer, move their
point of sale in a month. Ocado rebuilt into
the cloud in six hours. We used all of Evernote with
three petabytes of data in three months. That is how fast in the model
that I am describing, you can get there. So I think that
there was a speed limit. That speed limit was defined
by our ability to work with the data that you had and the
install base that you had and moving it into this new model. The new model works. Trust me. And if you're doing a new build,
you're clearly going to do that anyway because
you're intelligent. The problem was
you're stuck in between. You can't get to the new one. You're still
stuck on the old one. So I don't know. What are some examples? If you have data as an object,
we now have a software mechanism that will take the data and -
which we call by the way, simple transfer service, that will
basically just move it into our cloud structure and
just do it all over it. That is how the data
gets there so quickly. You built on containers. Again, we can support
every model of containers. You're built on Hadoop; we love
Hadoop in the sense that it is very similar to the
architectures that it was based on and it's easy to host it
inside of our model or use something even more
extensive like big query. You like building
virtual machines? As I said, just
move them straight over. So we've made huge strides in
terms of engineering investment and sales and customer service
investment to scale that. So how do you do this? Well in the first place, if you
leave the infrastructure to us, we put thirty billion dollars -
and I know because I approved it so it's real, right,
into this platform. Please do not attempt
to duplicate it. You have better
uses of your money. I would much rather have you
take the money that you have and the talents that you have and
build on top of this platform to provide that rapid iteration
and that rapid extension in the model that I am describing. What is interesting is that
we're continuing to take this stuff and make it stronger. We just launched Spanner and if
you don't know what Spanner is, Spanner is a - in my view,
a work of art in a computer science sense. It's a way of doing SQL, so it's
a proper database, but it's both globally essentially replicated
and also globally coherent at a scale of data that has
never been seen before. And by the way, we use it. It's how Google works. And by the way, we released
it a week ago for the cloud. We have a container builder
that does the same thing. So if you think about it, your
model is something like this. Big query provides petabyte
scale analysis on demand. Big table
provides millions of QPS. Google Cloud Storage
for the actual millions of objects, right? Again, the prices
are very, very low. And all of this platform that I
am describing, this get to the cloud now model, and most people
by the way are still in the moving their X86s into VMs,
getting some of the data out of the VMs, putting them into
proper databases, right? So you're still in the sort
of getting there phase. Once you have the data,
you have an opportunity for real transformation. I think that big data is so
powerful that nation states will fight over how much data
matters, right, that he who has the data, that can do the
analytics and the algorithms that Fei-Fei talked about at the
scale that we're talking about will provide huge nation state
benefits in terms of global companies and benefits for their
citizens and so forth and so on. But this is all basically
because once you have the data, you have to do one more thing. You have to change from writing
programs to instead building programs that learn outcomes. And what machine learning really
is and AI really is is it is a change to the way you program. And over and over again, and
we've talked about this to some length already as a company,
we've seen because of vision, right, we have the possibility
of drastically reducing the number of deaths in automobile
accidents and truck accidents, a big deal, right? A huge deal. And by the way, accidents in
cars are going up right now in America because distractors
appear to be - drivers appear to be distracted by the very
technology that is going into the cars. So we've got to solve that
problem and we've got to do it now. That is happening. In healthcare, in pathology,
radiology, all of these sort of patient-centered issues, we're
not that different from each other and the machine learning
and collaborative learning that is now possible really
will produce amazing healthcare outcomes. And there is a possibility that
the kind of intelligence that is going to be buildable on these
platforms will really surpass the kinds of insights that the
average person in a business is going to have. We're seeing over and over
again, people are using big data to do customer analysis, pattern
matching analysis, and customer targeting that really does
produce extra insight and that extra insight is worth billions
and billions of dollars in their marketing programs
and knowing it. So let me finish by saying
that we're here for real. You're here for real. This is an incredibly serious
mission, something I have wanted to do since I joined the
company seventeen years ago. It's something that I know that
the people you heard whenever they started, care a
great deal about it. The company has both the money,
the means, and the commitment to pull off a new platform of
computation globally for everybody who needs it, and
something which allows you to be satisfied that our principles
which are basically about openness and openness access
really will allow you both the freedom of choice - you won't be
locked in - but also the freedom to innovate. When I look at our partners and
I look at the speed with which they can now use their scarce
resources which are always programmers, to scale their
business solutions, their enterprise solutions, their
customer solutions, their video solutions, and their
marketing solutions, it just warms my heart. It's something we've looked
at - we've looked forward to for decades. With that, thank you very much
and I'm going to bring you back to Diane. Thank you. >> DIANE GREENE: So much. Eric - it's always so wonderful
to hear Eric's brilliant insights and perspectives. He was one of the original
enterprise people way back when. So I want to just recap what
you heard from our customers, our partners. You heard from Disney. They're doing a full lift
and shift to Google Cloud. They're using - they're
going after machine learning. I like to say bring
the magic to the magic. Our SAP partnership is so broad. You heard about certifying HANA. You heard about integrating with
G Suite, the machine learning partnership, and also this very
interesting partnership we're doing for a data custodian to
help our customers solve the difficult issues around
compliance, regulatory security. And then you heard from our
joint customer Colgate, a 211 year old company. You have to keep reinventing
yourself to make it that long and they are transforming
themselves today with G Suite. And then Verizon, moving 150,000
employees to G Suite, again going after a cultural
transformation and dealing with some really difficult issues
around moving to the cloud in their environment. Then we had Home Depot, very
progressive, forward looking retailer, looking to technology
to get the advantage, looking to machine learning to
get the advantage. Already taking advantage of
Google's scale to get them through things
like Black Friday. Then we had a very interesting
look into one of the world's largest banks, HSBC, about how
they're going to use Google Cloud and the technology and
the machine learning to really streamline their operations and
do things people didn't really realize was possible. And then we
closed out with EBay. They definitely see Google Cloud
as a technology that gives them huge competitive advantage. Very interesting how easily they
were able to run their catalog on our cloud and work
in a hybrid environment. Then of course we saw that
wonderful demo where the chat bot running on Google Cloud was
talking to Google Home, giving answers about the price of
a camera in their catalog. Just terrific stuff. I want to also you know -
where are all of the product announcements? All of the product announcements
- we have dozens of them. They're tomorrow when
you'll hear from our leaders. You'll hear from Urs Hölzle. You'll hear from Brian Stevens. You'll hear from Prabhakar
Raghavan around G Suite. You'll hear from Chet Kapoor,
the CEO of Apogee, telling you all kinds of new products that
we're bringing out and where we're going and
quite a bit of vision. Then Friday it's about
developers through and through. It's about open source,
startups, and Sam Ramji will lead that up. Before I wrap, I want to really
thank all of our partners, our precious partners, all
150 of them on the floor. I really look forward to running
into you on the showcase floor and personally, I also want to
express my deep appreciation to all of the thousands of
people in Google Cloud. I mean, you're amazing people. It's why I'm at Google. And the hard work you've done
over the last year, you know, and how smart and
excellent your work is. Much appreciation. I want to remind you - we're
going to take a break but at 12:30, we're going to have
two real internet giants. I was a judge in the first Queen
Elizabeth Prize for Engineering. I think it's now
in its third year. Vint Cerf and Marc Andreesen
were two of the winners for inventing the internet. Those two
haven't skipped a beat. They are actively leading
and defining our world and technology and
that's an ongoing basis. They'll be here for a
discussion with Quentin Hardy. So with that I'm
really wrapping it up. It's been a long keynote. Thank you for staying. We really appreciate it. And hey everybody. Enjoy the show. Thank you very much.
