Google Cloud Connect @ Korea
- Google Cloud Vision
- G Suite Product Roadmap
- Google Cloud Security
- Google Cloud Machine Learning
- G suite Customer Stories
2. Confidential + Proprietary
14:00 pm Keynote: Introducing Our Cloud Vision Chris Jang
Head of Google Cloud, Korea
14:20 pm Product Roadmap : What’s New in G Suite Kenneth Siow
Asia Head of Account Management
15:00 pm Introduction To Google Cloud Platform & Security
On The Cloud
Jerry Jeong
Sales Engineering Lead, Korea
15:30 pm Coffee Break
15:50 pm Driving Innovation with Google & Machine Learning Myunggeun Choi
Sales Engineering Lead, Korea
16:20 pm Customer Success Sharing with Megazone Estee Chi
Megazone
16:40 pm Closing
AGENDA
8. FASTER (US, JP, TW) 2016
Unity (US, JP)
2010
SJC (JP, HK, SG) 2013
Edge points of presence >100
Monet (US, BR) 2017
Network sea cable investments
PLCN Unity (HK, LA) 2018
Indigo (SG, ID, AU)
2019
Tannat (BR, UY, AR) 2017
Junior (Rio, Santos)
2017
Google global cache edge nodes (>800)
Google Global Network
Network
9. Google Cloud - Organizing Principles
Infrastructure will become a commodity
Rapid app development will drive innovation
Data is the new oil
AI & ML will change everything
10. Proprietary + Confidential
24%
discount
21%
discount
List price
differences
Per-minute billing,
custom machines,
sustained-use
discounts
100%
80%
60%
40%
20%
0%
Average 60%
advantage
가격 대비 성능 리더십
Rightsizing
recommen-
dations
15%
discount
Other Cloud
Providers
11. Proprietary + Confidential
오픈 소스
Co-innovation, interoperability,
and portability are attributes of
a future-proof architecture
Bring your own stacks,
including Cloud Foundry,
OpenShift, Docker, and Mesos
Languages Operating Systems Developer Tools Data Tools
12. Leading open source communities
#1Highest Engagement
on Github
#2
Highest Engagement
on Github
Kubernetes Tensorflow
Source: Analyzing GitHub issues and comments with BigQuery
15. Proprietary + Confidential
Fully-managed Kubernetes
container orchestration
Integrated platform for building
great mobile apps on iOS & Android
Convert on-prem VMs to
Compute Engine
Enterprise API
management & migration
Full Stack Support for
Windows, C#, .NET
Container Engine
Tools for app modernization & hybrid architecture
Compute Engine Stackdriver
16. Proprietary + Confidential
Data is the new oil
Clive Humby, 2006
We need to find it, extract it, refine it,
distribute it and monetize it.
“Data is the new oil” articles
Fortune, 2016
Wired, 2017
Economist, 2017
Inc, 2017
CNBC, 2017
CIO Magazine, 2017
17. Proprietary + Confidential
The Data Tsunami
“More data has been created in the last 2
years than in the entire history of the
human race.”
Data Drivers
Consumer internet; digital economy;
Internet of Things; machine data; etc.
18. Transform data into actions
Data
Preparation &
Processing
Cloud Dataflow
Cloud
Dataproc
Exploration &
Collaboration
Google
BigQuery
Cloud Datalab
Google
Analytics 360
Cloud
Dataproc
Mobile apps
Sensors
and
devices
Web apps
Developers
Data
scientists
Business
analysts
Data
Ingestion
Cloud Pub/Sub
App Engine
Databases/
Storage
Cloud SQL
Cloud Bigtable
Cloud
Datastore
Cloud Storage
Analytics
Google
BigQuery
Google
Analytics 360
Cloud
Dataproc
Google Drive
Advanced
Analytics &
Intelligence
Cloud Machine
Learning
Translate API
Vision API
Speech API
20. Build new 앱
and experiences
Make teams
productive
with secure
모바일/디바이스
Businesses are moving to
cloud for multiple purposes…
Optimize
인프라스트럭처
costs and
achieve scale
Gain value from
데이터 to predict
business outcomes
Connect to
비즈니스 플랫폼 of
services
and partners
25. G Suite:
Create, think, and reason together
through a connected experience.
G Suite:
From Productivity to Max Velocity
26. Attach & Email Comment Merge VersionsCreate
Repeat
Old habits are hard to break
G Suite:
From Productivity to Max Velocity
27. Access: Store files and find what you need. Instantly.
Control: Manage users, devices and data securely and easily.
G Suite:
From Productivity to Max Velocity
Create: Everything you need to bring your project to life.
Connect: Reach your colleagues, wherever they are.
30. G Suite
Basic
G Suite
Business
Intelligent Office Suite
G Suite
Enterprise
Email and Calendar
Docs, Sheets, Slides, Keep
Forms, Sites, Google+
Access across devices
Offline access
Open and edit Microsoft Office files
Ad-free experience
Google Cloud Search within G Suite
31. Google Cloud Search within G Suite
Google Search -- built for G Suite
● Cloud Search is just like Google Search -
built for G Suite. Search emails, docs,
events, the corporate directory and more
Get what you need, before you ask
● Open the Cloud Search app to get timely
and relevant assist cards that help keep
you organized throughout the day
Find people faster
● Search your company directory and see
colleagues’ contact details, plus events and
files you have in common
33. G Suite
Basic
G Suite
Business
Enterprise-class File Sync and Share
G Suite
Enterprise
Drive
Advance NLP search and Quick Access
IRM on Drive files
File versioning
Team Drives
Granular access and sharing controls
File audit reporting and alerts
Domain whitelisting
Storage per user**
Max file size
**1TB/user if fewer than 5 users
30GB Unlimited** Unlimited**
5TB 5TB 5TB
34. Team Drives
● Files owned collectively
○ Files are owned by the team so they don't
get lost if users leave
● Sharing permissions simple to manage
○ All team members see the same files and
a new edit access role prevents team
members from deleting files
● Easy onboarding of new team members
○ New team members automatically see
their team's content
35. Granular access and sharing controls
● Decide which departments or roles can sync files
and get offline access
● Customize file sharing by dept or role
● Restrict file sharing to trusted partners
● Secure all data with managed security keys
37. Domain Whitelisting
● Allow file sharing in Google Drive for organizations
that you trust by whitelisting their domain
example.co.jp
company.com
factory.com
Share files and
folders with users
in the trusted
domain
39. G Suite
Basic
G Suite
Business
Enterprise-class Meetings
G Suite
Enterprise
Hangouts
Instant access from mobile device
Easy guest access with link
Dial-in access to meetings
Video meetings (up to 25 participants)
Larger meetings (up to 30 participants)
Record meeting and save in Drive*
*Roadmap
40. Dial-in access to meetings
hangouts.google.com/hgw-ibeg
● Every meeting generates a short link and
dial-in phone number
● Up to 30 video participants
● Seamless integration into Calendar
● Instant screen sharing
43. G Suite
Basic
G Suite
Business
Device Management
G Suite
Enterprise
One click setup
Security policies for Android and
iOS BYOD devices
Android work profile with DLP controls
Enterprise app store
Monitor and audit entire mobile fleet
Automate device workflows
Bulk enrollment of company-owned
Android devices
45. Automate device workflows
Automate device
workflows
https://developers.google.com/admin-sdk/directory/v1/guides/manage-mobile-devices
https://developers.google.com/admin-sdk/reports/v1/guides/manage-audit-mobile
Time based work access
Access based on min OS/App version
Custom device approval workflow
Account Wipe after X failed unlocks
46. Bulk enrollment of company-owned Android devices
● Easy for admins to enroll all company owned devices
for MDM
● Easily enforce device-owner mode on multiple
Android company-owned devices
● Not be dependent on the end user to enable
device-owner mode
● gather info on each and every Android device/app
accessing G Suite
48. G Suite
Basic
G Suite
Business
Secure by Design
G Suite
Enterprise
Encryption in transit and at rest
99.9% uptime guarantee
Enterprise certifications and compliance
Vault for eDiscovery, archiving and
compliance
Advanced DLP for Gmail
Advanced DLP for Drive files
Security key enforcement
Third-party email archiving
Control 3rd-party data access with
app whitelisting
Encryption using your own S/MIME
certificates for Gmail
*Roadmap
Pre-delivery deep scanning of email
attachments*
49. Vault for eDiscovery, archiving and compliance
Archiving
Messages and on-the-record chats are securely managed
in-place, with no need to duplicate the data
Retention
Define retention policies to ensure business critical data is
preserved and properly managed
eDiscovery
Search, identify, preserve and collect data based on custodian,
content and metadata
50. Advanced DLP for Gmail and Drive files
2. Specify conditions to check for1. Set the Scope 3. Specify the appropriate action
Apply based on
● Domain or
● organizational unit
● Predefined content detectors
● Custom content detectors
● Detection thresholds
● Gmail
○ Reject emails
○ Quarantine emails
● Drive
○ Block external
sharing
○ Warn user
○ Send notification
To Admins
○ Audit sensitive files
51. Proprietary + Confidential
Security Keys for 2-Step Verification
Your Password
Security Key
Account Data
Based on FIDO U2F standard
● Helps protect against phishing
● Easy: Insert and press button
● Compact: One device, many services
53. Pre-delivery deep scanning of email attachments*
*Roadmap
● Pre-delivery scanning of email
attachments
● Opens attachments in a private, secure
sandbox to identify malicious files that
conventional scanning misses
● Neutralizes malware propagated through
embedded scripts
55. G Suite
Basic
G Suite
Business
Open and Extensible
G Suite
Enterprise
Integrate with apps in G Suite
Marketplace
APIs and Admin SDK for extending
G Suite functionality
App Maker for building custom apps*
Gmail Log Analysis in BigQuery
*Roadmap
56. G Suite Integration with BigQuery and Data Studio
Python hosted
Scripts on App
Engine
Data
Warehouse on
BigQuery
API Query &
Export
Data Load
Database
Query
3rd Party
Data
(ie: Demographic,
usage, etc)
Data
StudioG Suite
Admin SDK
57. Gmail Log Analysis in BigQuery
● Analyze gmail logs via sophisticated,
high performing custom queries
● Retain data for a configurable period
● Create custom reports and dashboards
using analytics tools such as Data
Studio, Tableau, etc
58. Example: G Suite recommends using gmail logs and BigQuery with Data
Studio to Understand employee behavior and activity. More here
Visualize with beautiful, informative reports
60. App Maker for building custom apps*
● Rapid visual application build & deploy
● Declarative data modeling
● Powered by Apps Script
● Designed for IT Developers
● No need to worry about compute,
storage, network
*Roadmap
63. Proprietary + Confidential
Check-in Kiosk app
“App Maker provides functionality which can
reduce the most time consuming part in building
webapps. PwC will be able to help our clients
create solutions that can be implemented with
less time through a collaborative use of App
Maker."
Brett Craven
Director, PwC Advisory, Google Premier Partner
64. Proprietary + Confidential
Vacation Requests
“AppMaker enables us to deliver in a short
timeframe valuable business applications that
perfectly extend the great features of the web
and Google G Suite specific.”
Richard van Delft
Managing Director, G-workplace, Google Cloud
Partner
65. Proprietary + Confidential
Weekly Sales Report
“App Maker really helped gPartner to gain deals
with costs reductions (30%) thanks to a
powerful interface with many pre-built tools as
the UX design or the integrated databases.”
Harold Ousset
Technical Lead, gPartner, Google Premier
Partner
66. Proprietary + Confidential
App Maker is great for
Add structure and control to ideas
and data
List management Dashboards Reporting
Process & Automation Alerts Utility
Keep up to date with connected
intelligence
Get visibility on what's most
important to your business
Turn the mundane and manual into
automatic and actionable
Don't miss what's most important Build apps as unique as your
business
72. Google Cloud Platform
Compute Big Data
BigQuery
Cloud
Dataflow
Cloud
Dataproc
Cloud
Datalab
Cloud
Pub/Sub
Genomics
Storage & Databases
Cloud
Storage
Cloud
Bigtable
Cloud
Datastore
Cloud SQL
Cloud
Spanner
Persistent
Disk
Machine Learning
Cloud Machine
Learning
Cloud
Vision API
Cloud
Speech API
Cloud Natural
Language API
Cloud
Translation
API
Cloud
Jobs API
Data
Studio
Cloud
Dataprep
Cloud Video
Intelligence
API
Advanced
Solutions Lab
Compute
Engine
App
Engine
Container
Engine
GPU
Cloud
Functions
Container-
Optimized OS
Identity & Security
Cloud IAM
Cloud Resource
Manager
Cloud Security
Scanner
Key
Management
Service
BeyondCorp
Data Loss
Prevention API
Identity-Aware
Proxy
Security Key
Enforcement
Internet of Things
Cloud IoT
Core
Data
Transfer
Appliance
73. Google Cloud Platform
Developer Tools
Cloud SDK
Cloud
Deployment
Manager
Cloud Source
Repositories
Cloud
Tools for
Android Studio
Cloud Tools
for IntelliJ
Cloud
Tools for
PowerShell
Cloud
Tools for
Visual Studio
Container
Registry
Google Plug-in
for Eclipse
Cloud Test
Lab
Networking
Virtual
Private Cloud
Cloud Load
Balancing
Cloud
CDN
Cloud
Interconnect
Cloud DNS
Cloud
Network
Cloud
External IP
Addresses
Cloud
Firewall Rules
Cloud
Routes
Cloud VPN
Management Tools
Stackdriver Monitoring Logging
Error
Reporting
Trace
Debugger
Cloud
Deployment
Manager
Cloud
Endpoints
Cloud
Console
Cloud
Shell
Cloud Mobile
App
Cloud
Billing API
Cloud
APIs
Cloud
Router
Cloud Partner
Interconnect
Container
Builder
79. DB Instance
Compute Engine
Single Instances
Persistent
Disk Snapshot
App Instance
Compute Engine
Single Instances
Cloud
DNS
1 App VM + 1 DB VM
● 분당 과금 ⇒ 초당 과금
● 자동 할인
● 누적 사용량 자동 합산
● 인스턴스 사이즈 조정
82. Instance Group
App Instance
Compute Engine
Multiple Instances
Instance Group
DB Instance
Compute Engine
Multiple Instances
Persistent
Disk Snapshot
Cloud Load
Balancing
Cloud
DNS
Scale Out to Multiple VMs
● Auto Scaling
● Instance Group
● Preemptible VM
(최대 80% 저렴)
86. Firebase Mobile Platform
Instance Group
App Instance
Compute Engine
Multiple Instances
Instance Group
DB Instance
Compute Engine
Multiple Instances
Persistent
Disk Snapshot
Cloud Load
Balancing
Cloud
DNS
Multiple
Devices
Native Apps or
Mobile Web
Firebase
90. 3단계 - 다양한 기능 추가
● 빠른 개발 및 배포
● 관리 부담 최소화
● 성능 확장
● 안정성
91. App Engine and Container Engine
Cloud
Storage
Cloud
SQL
Cloud
Datastore
Big
Table
Storage
Compute
Engine
(IaaS)
App
Engine
(PaaS)
Container
Engine
(powered by
Kubernetes)
Compute
BigQuery Pub/Sub Dataflow
Big Data
Dataproc Datalab Translate
API
Cloud
Vision
Speech
API
Machine Learning
Cloud ML
Firebase
92. Serverless & Containers at Scale
AppEngine Cloud FunctionsContainer Engine
Planet-scale
Platform-as-a-Service
Containerized app
management
Serverless apps,
Google infrastructure
93. Confidential + Proprietary
Powerful built-in services
Managed services, such as Task Queues, Memcache and the Users API, let you build any application.
Deploy at Google scale
You can scale up to 7 billion requests per day and automatically scale down when traffic subsides.
Focus on your code
Let Google worry about database administration, server configuration, sharding & load balancing.
Popular languages & frameworks
Write applications in some of the most popular programming languages, use existing frameworks and integrate
with other familiar technologies.
App
Engine
Familiar development tools
Use the tools you know, including Eclipse, IntellIJ, Maven, Git, Jenkins, PyCharm & more.
Multiple storage options
Choose the storage option you need: a traditional MySQL database using Cloud SQL, a schemaless NoSQL
datastore, or object storage using Cloud Storage.
94. Confidential + Proprietary
Container Engine - Kubernetes
● Container orchestrator
● Builds on containers (Docker, Rocket)
● Handles container and node failure
● Multiple cloud and bare-metal
environments
● 100% Open Source, written in Go
Let users manage applications, not machines
95. Confidential + Proprietary
Containers at Google
Google is using container
technology for more than 10
years
Everything at Google runs in
a container
We launch over 2 billion
new containers per week
96. Proprietary + Confidential
Cloud Functions
Cloud Functions respond to events that occur in your cloud infrastructure
Cloud services
emit "events"
1
Your Functions
respond to events
2
Functions can write
back to the cloud
Or call other APIs
3
함수만 작성하면 애플리케이션 완성...
97. Cloud Storage and Databases
Cloud
Storage
Cloud
SQL
Cloud
Datastore
Big
Table
Storage
Compute
Engine
(IaaS)
App
Engine
(PaaS)
Container
Engine
(powered by
Kubernetes)
Compute
BigQuery Pub/Sub Dataflow
Big Data
Dataproc Datalab Translate
API
Cloud
Vision
Speech
API
Machine Learning
Cloud ML
Firebase
98. Where do I store my data?
Cloud
Storage
Cloud
Bigtable
Cloud
Datastore
Cloud
SQL
Good for:
Binary or object
data (BLOB)
Such as:
Media, analytics,
archive/backup
Good for:
Hierarchical,
mobile, web
Such as:
User profiles,
Game State
Good for:
Web
frameworks
Such as:
CMS,
eCommerce
Good for:
Heavy read +
write, events,
Such as:
AdTech,
Financial, IoT
Cloud
Memorystore
Good for:
Web/mobile apps,
gaming
Such as:
Game state, user
sessions
EAP
Cloud
Spanner
Good for:
RDBMS+scale,
HA, HTAP
Such as:
User metadata,
Ad/Fin/MarTech
Big
Query
Good for:
Data
Warehouse
Such as:
Analytics,
Dashboards
In Memory Relational NoSQL Object Warehouse
Good for:
Local VM file
storage
Such as:
Application
data/binaries
Block
Persistent
Disk (GCE)
99. Google Cloud Storage
● 무한 확장 파일 저장소 (예: 상품 이미지, 로그, 문서)
● 오래 저장되고 억세스가 적어지는 파일은 자동으로 저비용 스토리지로 이동
Save cost by moving aging
data to Nearline.
(Seconds Latency)
Start with Multi-regional
or regional for highly
available, low-latency
data.
Move to Coldline as data
becomes needed less
frequently.
Age of your data over time
1 2 3
100. Cloud Spanner
글로벌 (리전간) 동기화 및 일관성 보장
표준 SQL 지원 (ANSI 2011)
주요 언어에 대한 SDK 지원
(Java, Python, Go, Node.js, etc.)
무한 확장이 가능한 RDBMS
NewSQL =
NoSQL의 횡적 확장성(Scale Out)과
RDBMS의 SQL과 테이블간 관계
(PK,FK)를 지원하며 트랜잭션에
대한 일관성을 보장하면서 글로벌
동기화를 지원하는 데이터베이스
101. Confidential & ProprietaryGoogle Cloud Platform 101
ACID Transaction + Query Language
Cloud Datastore
Battle-tested in App Engine for 7 years
Seamlessly scale with zero downtime
Fast and fully managed NoSQL
Serving 15+ Trillion Requests Per Month
102. Step 4 - 데이터 분석 (수익 극대화)
● 고객 유지
● 추천 엔진
● 마케팅 효과 극대화
● 재고 계획(예측)
103. BigData and Machine Learning
Cloud
Storage
Cloud
SQL
Cloud
Datastore
Big
Table
Storage
Compute
Engine
(IaaS)
App
Engine
(PaaS)
Container
Engine
(powered by
Kubernetes)
Compute
BigQuery Pub/Sub Dataflow
Big Data
Dataproc Datalab Translate
API
Cloud
Vision
Speech
API
Machine Learning
Cloud ML
Firebase
104. BigQuery for Data Analytics
● https://bigquery.cloud.google.com/
“ 1000억개(4.06TB) 의 문자열에 대해서 3개의 와일드 카드 연산을 수행 한후,
그룹핑 및 소팅을 하여 SUM을 구하는 연산 “
105. BigQuery for Data Analytics
비용
(4TB, 1000억개 레코드 분석에 20$)
스케일
(30초 동안 8600개의 CPU, 3600개의
디스크)
No-Ops
(별도의 운영이 필요없음)
클라우드 스케일의 성능
(1000억 레코드 4TB 분석에 30초)
Optimize your
time to answer
Programming
팀이 빅데이타 분석 및 개발에 집중
107. 커스텀 머신러닝 모델 Pre-trained 머신러닝 모델
Machine Learning
Engine
TensorFlow
Vision API
Translation
API
Natural
Language API
Speech API Jobs API
두 가지 타입의 머신 러닝
Video
Intelligence API
108. G-Suite and Google Cloud Platform
Infrastructure
(IaaS)
Data &
Analytics
AppDev
(PaaS)
Productivity
(SaaS)
Identity &
Security
Machine
Learning
Management
Tools
GCP의
계정 및 조직
관리
110. Google-Grade Security is Yours
Full Stack Ownership
● Top 5 server
manufacturer
● We build everything.
Hardware + custom Linux
software stack
+ networking
● Homogeneity and
ownership of full
stack greatly reduces
security footprint
Compliance
● ISO 27001
● SOC 1, 2 & 3
● HIPAA (BAA)
● PCI
Physical
● Remotely located and
restricted data centers
● Guards, cameras,
and advanced video
analytics to detect
intruders
● “Mission Impossible”
tech: Thermal imaging
cameras, perimeter
fences, laser detection,
and biometrics scans
People
● 450 top experts
● Discovered Heartbleed
● First to start reward
program for reporting
vulnerabilities
111. Google Cloud Platform 111
인증 보안 강화
● Multi-Factor Authentication
● Root 계정 및 텍스트 암호를 통한 인증 비활성화
112. Google Cloud Platform 112
최소한의 필요한 권한만 부여
● 리소스(서비스) 별 권한 부여
● 슈퍼 관리자 지양 ⇒ 업무 별 관리자
● 최소한의 권한 부여 후, 요청이 있을 때 추가 권한 부여
113. Google Cloud Platform 113
User 대신 Group에 권한 부여
● 유연한 권한 관리 가능
● 일관된(consistent) 권한 가능 (예) Admin들에게는 동일한 권한
114. Google Cloud Platform 114
Audit Log
● 권한 침해 문제 발생을 빠르게 발견 및 정확한 분석 가능
● GCP Logging 서비스 이용
○ Stackdriver Agent를 이용하여 중앙에 수집
● 로깅 대상
○ 리소스 접근 시도 및 결과
○ IAM Policy 변경
○ 애플리케이션에 의한 액세스 포함
115. Google Cloud Platform 115
서비스 계정 또는 SSH Key Rotate
● Key 유출로 인한 피해 최소화
● Google Managed Key의 경우 GCP에 의해 자동 Rotation
● User Managed Key는 API를 사용하여 자동 Rotation 구현
● API를 사용하여 SSH Key
자동 Rotation 구현
124. Driving Innovation with Google &
Machine Learning
최명근
Korea Sales Engineering, Cloud Platform
125. Agenda
01 The What/Why/How of Machine Learning
02 Leveraging GCP for Machine Learning
03 Customer Success Stories
126. Machine learning will be the
basis and fundamentals of
every successful huge IPO
win in 5 years.
— Eric Schmidt , Google NEXT 2016
“
”
127. Artificial Intelligence, deep
learning, machine
learning — whatever you’re doing if
you don’t understand it — learn it.
Because otherwise you’re going to
be a dinosaur within 3 years.
—Mark Cuban, Upfront Summit 2017
“
”
132. How does a
machine learn?
PREDICTING HOUSE SALE PRICES
Sq. Footage Price
1,000 $100,000
3,000 $300,000
How much would a 2,000 sq ft. house sell for?
133. How does a
machine learn?
PREDICTING HOUSE SALE PRICES
Sq. Footage Price
1,000 $100,000
3,000 $300,000
2,000 $200,000
134. 1
Look at the Data
Sq. Footage
Price
1,000 2,000 3,000
100,000
200,000
300,000
137. A better dataset
PREDICTING HOUSE SALE PRICES
Sq.
Footage
Crime Rate
(1-100)
School
Rating
(1-5)
Price
1,000 1 5 $300,000
2,000 50 3 $200,000
4,000 80 1 $100,000
How much would a 3,000 sq ft. house with a
crime rate of 30 and a school rating of 2 cost?
142. Largest Machine Learning repository on GitHub
Operates over tensors: n-dimensional arrays
Using a flow graph: data flow computation framework
A brief look at TensorFlow
● Train on CPUs, GPUs
● Run wherever you like (local, cloud, mobile)
143. WHAT
Algorithms that can generate insights by
learning from data.
WHY
Because algorithms can learn faster, cheaper,
and better than humans.
HOW
By finding patterns in data.
Review:
The What/Why/How
of Machine Learning
144. Agenda
01 The What/Why/How of Machine Learning
02 Leveraging GCP for Machine Learning
03 Customer Success Stories
151. Custom ML models Pre-trained ML models
Machine Learning
Engine
TensorFlow
Vision API
Translation
API
Natural
Language API
Speech API Jobs API
Two Flavors of Machine Learning
Video
Intelligence API
152. ● State of the art machine learning
models built by Google experts
● Pre-Trained on massive datasets
using Google scale infrastructure
● Work out of the box. No machine
learning expertise required
Machine
Learning APIs
153. Vision API
Detect broad sets of
categories within an image,
ranging from modes of
transportation to animals.
Analyze facial features to
detect emotions: joy,
sorrow, anger.
Detect logos.
Detect and extract text
within an image, with
support for a broad range of
languages, along with
support for automatic
language identification.
Extract text
Detect different types of
inappropriate content from
adult to violent content.
Powered by Google Safe
Search
Detect inappropriate contentObject Recognition Facial sentiment & logos
TRY THE API
154. VISION API
var vision = require('google-vision-api-client');
var requtil = vision.requtil;
//Prepare your service account from trust preview certificated project
var jsonfile = '/Users/terry/dev/ws/nodejs/GoogleVisionAPISample/My Project-eee0a2d4532a.json';
//Initialize the api
vision.init(jsonfile);
//Build the request payloads
var d = requtil.createRequests().addRequest(
requtil.createRequest('/Users/terry/images/dale2.jpg')
.withFeature('FACE_DETECTION', 3)
.withFeature('LABEL_DETECTION', 2)
.build());
//Do query to the api server
vision.query(d, function(e, r, d){
if(e) console.log('ERROR:', e);
console.log(JSON.stringify(d));
});
22 LINE OF
SOURCE CODE
(node.js face detection, label detection)
156. Natural Language API
Identify entities and label by
types such as person,
organization, location,
events, products and media.
Enables you to easily
analyze text in multiple
languages including
English, Spanish and
Japanese.
Extract tokens and
sentences, identify parts of
speech (PoS) and create
dependency parse trees for
each sentence.
Syntax analysisEntity Recognition Multi-Language Support
TRY THE API
Understand the overall
sentiment expressed in a
block of text.
Sentiment Analysis
157. Speech API
Powered by deep
learning neural
networking to power
your applications..
No need for signal
processing or noise
cancellation before
calling API. Can
handle noisy audio
from a variety of
environments.
Noise Robustness
Can provide context
hints for improved
accuracy. Especially
useful for device and
app use cases.
Word HintsSpeech Recognition
TRY THE API
Recognizes over 80
languages & variants.
Can also filter
inappropriate content
in text results
Over 80 languages
Can stream text
results, returning
partial recognition
results as they
become available.
Can also be run on
buffered or archived
audio files.
Real-time results
159. Translation API
Supports more than 100
languages and thousands
of language pairs.
Behind the scenes,
Translation API is learning
from logs analysis and
human translation
examples. Existing
language pairs improve and
new language pairs come
online at no additional cost.
Sometimes you don’t know
your source text language in
advance. Can automatically
identify languages with high
accuracy.
Automatic language
detection
The Premium edition is
tailored for users who need
precise, long-form
translation services (e.g.
livestream translations, high
volume of emails, detailed
articles and documents)
Premium edition BETA
Text Translation Continuous Updates
TRY THE API
160.
161. BETA: Video Intelligence API
Detect entities within the
video, such as "dog",
"flower" or "car".
You can now search your
video catalog the same way
you search text
documents..
Extract actionable insights
from video files without
requiring any machine
learning or computer vision
knowledge.
Enable Video Search
More features will be added
to the Video Intelligence API
during the BETA period.
More to come ... BETA
Label Detection Insights From Videos
167. Custom ML models Pre-trained ML models
Machine Learning
Engine
TensorFlow
Vision API
Translation
API
Natural
Language API
Speech API Jobs API
Two Flavors of Machine Learning
Video
Intelligence API
168. As of 2-17-2017
Most Popular
Machine Learning
Framework
Framework
GitHubStars
TensorFlow
10,000
20,000
30,000
0
40,000
SKLearn Caffe CNTK MXnet Torch
50,000
169. TensorFlow
● Tensorflow is what we use for our own
internal machine learning projects, and
now it’s available to you!
● Powers over 100 Google services
● Tutorials to get started at
https://www.tensorflow.org
170. Agenda
01 The What/Why/How of Machine Learning
02 Leveraging GCP for Machine Learning
03 Customer Success Stories
176. 액션 아이템
구글 클라우드 무료 사용 등록 및 쿼터 확인
cloud.google.com/free
구글 클라우드 플랫폼 블로그
https://cloudplatform.googleblog.com/
구글 클라우드 플랫폼 사용자 그룹
www.facebook.com/groups/googlecloudkorea
구글 클라우드 코드랩
https://codelabs.developers.google.com/
178. G Suite 을 선택한 기업들
( 메가존 고객 사례 중심으로 )
메가존 / Platform Service / 지연숙
179. 100여 가지의 의약품과 의약외품 등을 생산 판매하고 있고, 신약 연구개발도 활발히 진행하고 있는 규모 약 800
여명의 국내 성공적인 제약 회사. 일반 약품으로 잘 알려진, 아락실, 시린메드 치약, 부광 안티프라그 등이 있다.
[ 도입 배경 ]
● 로터스 노츠 서버의 모바일 접근성 및 기업 정보의 접근성 한계
● 다른 경쟁사 브랜드는 비용면에서 사업 요구 사항 충족하지 못했음
● 최고 경영자 및 임원진의 강력한 서포트
[ 결과 ]
● 어디서든지 모바일 기기를 통해서 이메일과 내부 문서 접근 가능
● 사무실이 아니더라도 메일에 답변할 수 있고, 드라이브 및 각종
도구를 활용하여 시장조사 결과를 실시간으로 보고하여 업무 효율
향상
“우리 직원들은 특히 Google+가 가져다 준 증진된 협업과
관련된 혜택을 누리고 있습니다. Google Apps는 직원들이
어디에서나 근무할 수 있도록 해주는 비용 효율적인 솔루션을
제공함으로써 우리의 사업과 관련된 요구사항을
충족시켜줍니다.“
—부광약품(주) 개발팀 이정욱 상무
180. SK 플래닛은 변화와 혁신으로 세계 시장을 선도하는 Integrated & Personalized Commerce 전문기업 입니다.
Commerce Platform 으로는 11번가, Marketing Platform 으로는 OK 캐쉬백, Syrup 등이 있습니다.
[도입 배경]
● 팀별로 Consumer 용 Gmail 및 Drive 사용이 늘어남
● 회사에 귀속될 수 있는 계정 및 데이터 소유권 문제 대두
● 팀간, 부서간 신속한 협업 도구에 대한 니즈
[ 특징 ]
● Drive 및 문서 도구, 행아웃 사용
181. 및 계열사 (플레로 게임즈, 조이맥스, 위메이드 아이오)
[도입 배경]
● MS O365 One Drive 속도 (업로드, 다운로드 속도) 이슈
● 내부 탈 MS 움직임
● 가격대비 G Suite 기능, 안정성, 보안, 확장에 대한 만족
[ 특징 ]
● 아웃룩 클라이언트 사용에 대한 의존도가 높았음
● 브라우져 기반의 Gmail UI 에 대한 생소함으로 인해 변화 관리,
사용자 교육을 집중적으로 진행
● 최근 6개월 전사적으로 공유한 파일 수 변화 패턴 (아래 참조)
온라인 / 모바일 게임 개발을 진행해 왔으며, 2009년 12월에 코스닥에 상장되었고 미르의 전설2, 미르의 전설3
등 MMORPG 를 중심으로 많은 고객군을 확보하고 있다.
182. 부동산 정보 전문앱 No.1
[ 도입 배경 ]
● 국내 호스팅 메일 서비스 사용중 이었으나, 급속하게
성장하는 기업의 비즈니스 니즈에 걸맞는 글로벌 솔루션
니즈
● 단순한 메일 서비스가 아닌, 협업의 가치 및 구글의
일하는 방식에 대한 임팩
[ 특징 ]
● 클라우드 및 SaaS 에 개방적인 조직 문화
● 현재 니즈를 해소할 수 있는 솔루션인지 검증에 집중
● 활발한 전사 활용 : 메일, 문서공유, 캘린더 등
183. F&F (Discovery, Banila Co, MLB 등)
아웃도어 Discovery, 화장품 브랜드 Banila Co, 아동 명품 브랜드 MLB 제품을 생산하는 의류 제조 및 유통 전문
업체.
[ 도입 배경 ]
● 해외 G Suite 사용 사례 임팩
● 기존 이원화된 시스템 통합
● SaaS 기반 솔루션 (비용, 운영, 혁신)
● G+, Hangout Chat, 문서공유
[ 현재 상황 ]
● 사용자 활용도를 높이기 위한 변화관리
● 운영 및 지속적인 혁신을 위한 서포트
[ 특징 ]
● 전국 매장과의 협업 및 소통
● 국내 대형 패션 유통 첫 사례
184. ● 네트워크 환경이 SaaS 사용에 결정적 장애가 될만한 요소가 있는가? 를 체크해야 한다.
● 내부 보안 정책과 배치되는 사항이 있는지 점검해야 한다. (필요하다면, 충분한 검토후
정책의 방향성을 변경하는 고려도 할 수 있어야 한다.)
● 작고 사소해보이는 기능 (G Suite 에 없는) VS 도입으로 얻게되는 큰 변화에 대한 득실을
계산해 봐야 한다.
○ ex) 조직도가 꼭 필요한가? 우리는 왜 조직도를 사용해 왔는가?
● IT 주도의 도입이라면 반드시 비즈니스 사이드 임원진 후원 필요 필요하다.
● 처음에는 모든 도구를 다 사용하지 않더라도, 단계별 변화관리를 통해 활용범위를 넓힐
계획을 세워야 한다.
● 단순한 이메일 시스템이 아닌만큼, 솔루션이 가진 가치를 최대한 활용하려고 해야한다.
● 지속적으로 발표되는 G Suite 추가되는 기능들, 기술의 동향을 활용해야한다.
G Suite 도입시 고려해야할 사항 (권고)