Solving Business Problems
with
Machine Learning
Google Analytics Conference
Vienna April 19 2018
Lukman Ramsey
lramsey@google.com
Head of AI Solutions
Google Cloud
Democratizing AI and Machine Learning
Three flavors of Machine Learning
Machine Learning APIs
Building custom ML solutions on Google Cloud
Customer Success Stories
Agenda
© 2017 Google Inc. All rights reserved.© 2018 Google LLC. All rights reserved.
Democratizing AI and ML
AI?Artificial
Intelligence
Machine Learning is...One branch of the field of
Artificial Intelligence
Underneath it all is
machine learning
data algorithmscomputations
Two Models Of Computation
Turing Machines
Von-Neumann Architecture
Algorithms Programmed by Humans
Symbolic Vector Space
Brain Machines
Biologically Inspired (Evolutionarily Evolved) Architecture
Algorithms Learned from Experience
Confidential & Proprietary
Source: Data scientists= Kaggle Data scientist community , Developers: Evans Data Corporation the figure in 2016 was 21m
State of the Industry: Lack of Expertise
Very few users today
can create a custom ML model.
To democratize AI,
we need to make AI accessible
to millions more
1000’s
Deep Learning
Researchers
21M
Developers
Confidential & Proprietary
<1M
Data Scientists
Confidential & ProprietaryConfidential & Proprietary
UPDATEDEPLOYEVALUATETUNE ML MODEL
PARAMETERS
ML MODEL DESIGN
DATA
PREPROCESSING
State of the Industry: Complex & Time Intensive
Large computational resource . Machine learning expertise . Manual data labeling
Google is an AI company
Confidential & ProprietaryGoogle Cloud Platform 11
Rapidly accelerating use of deep learning at Google
Google Cloud Platform 11
Google directories containing Brain Models
2012 2013 2014 2015
3000
2000
1000
0
Used across products:
4000
2016
Uniqueprojectdirectories
2017
Confidential & Proprietary
Democratize AI by making it accessible,
fast and useful for enterprises and developers
© 2017 Google Inc. All rights reserved.© 2018 Google LLC. All rights reserved.
Three Flavors of Machine
Learning
Proprietary + Confidential
ML Frameworks:
Total Control for Power
Users
Machine Learning APIs:
Ready to Go
Cloud
Vision API
Cloud
Translation API
Cloud Natural
Language API
Cloud
Speech API
Cloud ML EngineTensorFlow
Cloud
Jobs API
Cloud Video
Intelligence API
AutoML:
Bring Your Own Data
(We Do the Rest)
Pick Your Flavor
Cloud DataprocSpark ML
Proprietary + Confidential
Ease of Use vs Flexibility
Proprietary + Confidential
ML Frameworks:
Total Control for Power
Users
Machine Learning APIs:
Ready to Go
Cloud
Vision API
Cloud
Translation API
Cloud Natural
Language API
Cloud
Speech API
Cloud ML EngineTensorFlow
Cloud
Jobs API
Cloud Video
Intelligence API
AutoML:
Bring Your Own Data
(We Do the Rest)
Pick Your Flavor
Cloud DataprocSpark ML
ML APIs
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
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
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
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
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
@glaforge @manekinekko
Cross-platform tool for building advanced
conversational interfaces
ListSessions(
topics.ComputeEngine);
“What Compute Engine
talks are there?”
@glaforge @manekinekko
Integrate with...
Actions on Google
● Google Home, Pixel…
● and more to come
External integrations
● Slack, Facebook Messenger,
● Twitter, Twilio, Skype, Tropo,
● Telegram, Kik, LINE, Cisco Spark,
● Alexa, Cortana
Demo: Vision API
Demo: Natural Language API
Proprietary + Confidential
ML Frameworks:
Total Control for Power
Users
Machine Learning APIs:
Ready to Go
Cloud
Vision API
Cloud
Translation API
Cloud Natural
Language API
Cloud
Speech API
Cloud ML EngineTensorFlow
Cloud
Jobs API
Cloud Video
Intelligence API
AutoML:
Bring Your Own Data
(We Do the Rest)
Pick Your Flavor
Cloud DataprocSpark ML
AutoML
Confidential & ProprietaryGoogle Cloud Platform 29
AI expertise + Data
+ Computation
We call this
AutoML
Current solution But we can turn this into
Data + 100x Computation
How does AutoML work?
Controller: proposes ML models Train & evaluate models
20K
times
Iterate to
find the
most
accurate
model
Confidential & ProprietaryGoogle Cloud Platform 31
AI does AI
Systematic exploration
of the model space, using
the techniques finessed in
AlphaGo, yields super-human
performance in AI network design
CIFAR-10 Image Recognition Task
AutoML for Cloud Customers
Dataset Baseline AutoML
Customer 1 (Media) 75% 99%
Customer 2 (Housing) 87% 93%
Customer 3 (Wildlife) 85% 95%
Customer 4 (Sports) 90% 96%
Customer 5 (Insurance) 0.7 mean AUC 0.95 mean AUC
Results for AutoML on image problems https://cloud.google.com/automl/
Proprietary + Confidential
ML Frameworks:
Total Control for Power
Users
Machine Learning APIs:
Ready to Go
Cloud
Vision API
Cloud
Translation API
Cloud Natural
Language API
Cloud
Speech API
Cloud ML EngineTensorFlow
Cloud
Jobs API
Cloud Video
Intelligence API
AutoML:
Bring Your Own Data
(We Do the Rest)
Pick Your Flavor
Cloud DataprocSpark ML
Core technology for ML
● PaaS for Tensorflow
● Instantly scale your training up to 100 workers
(industry leading)
● Automatic monitoring and logging
● Seamlessly transition from training
to prediction
● Built in model version management
● No lock-in. Option to download your trained
models for on-premise or mobile deployment
Cloud ML Engine
Automatically tune your model with HyperTune
● Automatic hyperparameter tuning service
● Build better performing models faster and save
many hours of manual tuning
● Google-developed search algorithm efficiently
finds better hyperparameters for your
model/dataset
● Flexible: Hyperparameters are provided to to
user code as command line flags, allowing any
post-processing you want.
HyperParam #1
Objective
Want to find this
Not these
HyperParam
#2
TensorFlow
Google Use Of TensorFlow: # of Models
Search
Gmail
Translate
Maps
Android
Photos
Speech
YouTube
Play
… many others ...
Production use in many areas:
Research use for:
100s of projects and papers
Internal TensorFlow launch
Google-designed custom ASIC built
and optimized for TensorFlow
1st generation used in production for
over 16 months
Now on 2nd generation—180 Teraflops
per TPU
TensorFlow Research Cloud—1000
TPUs for researchers, at no charge.
Tensor Processing Unit
TPU Pod
64 2nd-gen TPUs
11.5 petaflops
4 terabytes of memory
2-D toroidal mesh network
© 2017 Google Inc. All rights reserved.© 2018 Google LLC. All rights reserved.
Building custom ML solutions on
Google Cloud
Proprietary + Confidential
Define ML use cases
Define specific ML use cases
for the project
Select algorithm
Choose the right ML
algorithm for the task
Build ML model
Develop the first iteration
of the ML model
Present results
Present results of the model in
a way that demonstrates its
value to stakeholders
Iterate ML model
Refine the ML model to
improve performance and
efficacy
Data pipeline &
feature engineering
Create the right features
from raw data for the
ML task
Plan for deployment
Prepare for deployment in
production
Operationalize model
Deploy and operationalize
ML model in production
Monitor model
Monitor deployed ML model
and retrain or rebuild when
performance degrades
1 3
10 789
Data exploration
Perform exploratory data
analysis to understand the
data
2 4
6
5
Start
a new ML project
with PSO
Cloud Discover Cloud MVM Cloud Deploy
Machine Learning Lifecycle
Building an ML model requires 3 things
Data
Compute
Talent
Data Scientist
Software Engineer
Data
Data
Compute
Talent
Data Scientist
Software Engineer
It's not who has the best
algorithm who wins, it's who
has the most data.
— Andrew Ng , Co-Founder of Google Brain
“
”
Capture
Pub/Sub
Process
Dataflow
Dataproc
Store
Cloud Storage
BigQuery
Cloud SQL
Datastore
BigTable
Analyze
BigQuery
Dataflow
Datalab
ML starts with getting a handle on your data
Insight
Cloud ML Engine
● Easily access and analyze
public and commercial
datasets hosted on GCP
Commercial Datasets
Program
Compute
Data
Compute
Talent
Data Scientist
Software Engineer
Talent
Data
Compute
Talent
Data Scientist
Software Engineer
Talent
● Every organization has people
capable of building ML
systems
● But those people may not
have the training and tools
they need to be successful
with machine learning
● Google provides both
Training
● Google Professional Services
will bring Google Machine
Learning expertise to your
company
● Intensive trainings and
workshops from 1 to 4 weeks
● Customized to your needs
Talent
Proprietary + Confidential
We Can Help You Implement your Solution
Google PSO1 Google + Partner2 3
Implement
Machine
Learning
APIs
Build
ML Models
Cloud
ML Engine
Deploy and ManageAnalyze and Plan
Google Cloud
Google Cloud
Proprietary + Confidential
Define ML use cases
Define specific ML use cases
for the project
Select algorithm
Choose the right ML
algorithm for the task
Build ML model
Develop the first iteration
of the ML model
Present results
Present results of the model in
a way that demonstrates its
value to stakeholders
Iterate ML model
Refine the ML model to
improve performance and
efficacy
Data pipeline &
feature engineering
Create the right features
from raw data for the
ML task
Plan for deployment
Prepare for deployment in
production
Operationalize model
Deploy and operationalize
ML model in production
Monitor model
Monitor deployed ML model
and retrain or rebuild when
performance degrades
1 3
10 789
Data exploration
Perform exploratory data
analysis to understand the
data
2 4
6
5
Start
a new ML project
with PSO
Cloud Discover Cloud MVM Cloud Deploy
Machine Learning Lifecycle
© 2018 Google LLC. All rights reserved.
2-day workshop, up to 5 days
Objectives
● Gain new competitive advantages with ML
● Identify potential ML use cases
● Address targeted business problems
● Survey the foundation for ML potential
Activities and Deliverables
1. ML overview session
2. Use case ideation workshop
3. High-level data qualification
4. Analysis and recommendations
Cloud Discover:
Machine Learning
Cloud Discover: Machine Learning helps
you understand machine learning (ML)
concepts and identify and qualify
potential business problems that can be
addressed using ML.
Ideal for determining if ML is right for
your business and which use cases are
realistic and achievable.
© 2018 Google LLC. All rights reserved.
2+ months
Engagement activities
● Develop an ML solution model
● Perform exploratory data analysis
● Explore the right set of features to include
● Present results of the model that shows its
business value
Deliverables
1. Data exploration
2. Data pipeline and feature engineering
3. Build and iterate ML model
4. Present results to stakeholders
Cloud Deploy ML
for MVM
Cloud Deploy ML for MVM (minimum
viable model) helps you take the use
case identified in Cloud Discover from
theoretical to practical by developing an
ML model.
This model will prove the value of the
use case and its ML solution to
stakeholders prior to investing more in
the next phase of build out.
© 2018 Google LLC. All rights reserved.
Cloud Deploy ML for MVM details
Implements steps 2–7 in the ML lifecycle
Delivers a minimum viable ML model or feasibility study
Delivers 32 days of work effort in a 2-month
calendar window
Prices start at $160K for 1 SCE at 100% and 1
consultant at 50%
© 2018 Google LLC. All rights reserved.
● Data exploration
● Algorithm selection
● Data pipeline
● Feature engineering
● Development of initial ML model
● Iteration to improve performance of ML model
● Building a complete data pipeline
● Deploying the model into production
● Converting the model into an API
In scope:
Out of scope:
Cloud Deploy ML for MVM scope
© 2018 Google LLC. All rights reserved.
How do we deploy?
(detailed level)
32-day work effort (minimum)
Advisory on ML model
deployment
Prerequisite: Discover ML
Typical machine learning journey
Assess Accelerate Transform
Cloud Discover
Assessment
What do we? Can we?
7-day FTE work effort
ML concepts training
Group ideation with
business and data
owners of use cases
Use case technical
qualification
Next steps on how to
create a test model for
the use cases
•
•
•
•
•
•
Step 1 should be taken to determine
a feasible ML use case. Qualified
customers can proceed directly to
Step 3 unless they need training or
are not fully committed to starting a
formal ML project.
Data Engineering
/GCP class
What is ML? How to
use data?
4-day training w/
hands-on lab
•
•
ASL: Immersion
Education
How to code ML?
1-month immersive
training at Google using
public datasets
•
•
Cloud Deploy ML
for MVM
•
•
•
•
ASL: Solution
Development
Let’s build an ML model together
6-month solution development
w/Google engineers for one use
case
Prerequisite: qualified use case
•
•
•
Cloud Deploy ML
for Production
How do we deploy?
(detailed level)
20-day work effort
(minimum)
Advisory on ML model
deployment
Prerequisite: ML MVM
•
•
•
•
Packaged Solutions
Confidential & Proprietary
Production Recommendation Solution on GCP
Google Analytics
BigQuery
Google Analytics
360
Customer Web
Application
Web
Server
Application
Server
Database
Server
Rec API
App Engine
Cloud Endpoints
Model Training
Cloud Machine Learning
Orchestration
Cloud Composer
ML Data
Training
Model files
Browser
Client
Mobile /
Tablet Client
Confidential & Proprietary
Kurier.at
3rd Largest news provider in Austria
Parent company: 500 web site properties
Google Analytics 360 customer
Outbrain user… Not happy
eDialog
GA Partner
© 2017 Google Inc. All rights reserved.© 2018 Google LLC. All rights reserved.
Customer Success Stories
Google Cloud Customers: Data & Machine Learning
© 2018 Google LLC. All rights reserved.
• Risk analytics and regulation
• Fraud detection
• Credit worthiness evaluation
• Customer segmentation
• Cross-selling and upselling
• Sales and marketing campaign
management
Financial Services
• Aircraft scheduling
• Dynamic pricing
• Social media – consumer feedback
and interaction analysis
• Customer complaint resolution
• Traffic patterns and congestion
management
Travel and Hospitality
Explore machine learning use cases with Google
• Predictive inventory planning
• Recommendation engines
• Upsell and cross-channel marketing
• Market segmentation and targeting
• Customer ROI and lifetime value
Retail
• Alerts and diagnostics from real-time
patient data
• Disease identification and risk stratification
• Patient triage optimization
• Proactive health management
• Healthcare provider sentiment analysis
Healthcare and Life Sciences
• Predictive maintenance or
condition monitoring
• Warranty reserve estimation
• Propensity to buy
• Demand forecasting
• Process optimization
• Telematics
Manufacturing
• Power usage analytics
• Seismic data processing
• Carbon emissions and trading
• Customer-specific pricing
• Smart grid management
• Energy demand and supply optimization
Energy, Feedstock and
Utilities
complex
solve
The ability to detect patterns in satellite images — such as the
difference between snow and clouds —​is critical to Airbus Defense
and Space’s users who depend on highly precise, up-to-date and
reliable information.
Produits Utilisés
Google Cloud Dataflow, BigQuery, Cloud Storage et Cloud Datalab
Industry: Aerospace – Region: France
“In our tests, Google Cloud Machine Learning enabled us
to improve the accuracy and speed at which we analyze
the images captured from our satellites. It solved a
problem that has existed for decades" .
Mathias Ortner, Data Analysis and Image Processing Lead
Solutions
One of our customers, Airbus Defense and Space, tested the use of
Google Cloud Machine Learning to automate the process of
detecting and correcting satellite images that contain imperfections
such as the presence of cloud formations.
problems
opportunities
identify
with advancements such as
Machine Learning
Speed
at which we analyze the images
captured
Accuracy
improved thanks to Machine
Learning
Gain
Insights
from images by detecting
individual objects and concepts
Intelligent structure for
30 billion
files managed with powerful
capabilities
Speed up
search and discovery with
image-centric workflows
Enabled accuracy for automation
performance
Industry: Technology – Region: North America
Box customers unlock new value from content as they
automatically classify images and greatly accelerate
business processes.
Improved
Extensive
content management for
customers in every industry
Products Used
Google Cloud Vision API
Solution
Google Cloud Platform and machine learning enable Box to help its
customers manage and gain insight from their image files, and
speed up image-centric processes and workflows.
119languages and dialects
supported
Machine learning that
Enables
applications that react to
what people say
Understand
intent of callers in addition to what
they say
Cutting-edge speech recognition
capabilities
Products Used
Google Cloud Speech API
Industry: Technology – Region: North America
“Machine Learning has changed the game.”
Jeff Lawson, CEO, Twilio
Expanded
support in Twilio Understand
with Google Cloud Speech API
Solution
Google Speech API enables Twilio to make it easier for developers
to build applications that react to what people say during phone
calls, taking callers speech and turn it to text. By adding a layer of
machine intelligence over existing support, customers can bypass
navigating menus and using phone keypads.
Machine learning is a core,
transformative way by which we’re
rethinking how we’re doing
everything.
- Sundar Pichai
“
”
Thank you
lramsey@google.com
Proprietary + Confidential

Google Analytics Konferenz 2018_Machine Learning / AI mit Google_Lukman Ramsey (Google US)

  • 1.
    Solving Business Problems with MachineLearning Google Analytics Conference Vienna April 19 2018
  • 2.
  • 3.
    Democratizing AI andMachine Learning Three flavors of Machine Learning Machine Learning APIs Building custom ML solutions on Google Cloud Customer Success Stories Agenda
  • 4.
    © 2017 GoogleInc. All rights reserved.© 2018 Google LLC. All rights reserved. Democratizing AI and ML
  • 5.
    AI?Artificial Intelligence Machine Learning is...Onebranch of the field of Artificial Intelligence
  • 6.
    Underneath it allis machine learning data algorithmscomputations
  • 7.
    Two Models OfComputation Turing Machines Von-Neumann Architecture Algorithms Programmed by Humans Symbolic Vector Space Brain Machines Biologically Inspired (Evolutionarily Evolved) Architecture Algorithms Learned from Experience
  • 8.
    Confidential & Proprietary Source:Data scientists= Kaggle Data scientist community , Developers: Evans Data Corporation the figure in 2016 was 21m State of the Industry: Lack of Expertise Very few users today can create a custom ML model. To democratize AI, we need to make AI accessible to millions more 1000’s Deep Learning Researchers 21M Developers Confidential & Proprietary <1M Data Scientists
  • 9.
    Confidential & ProprietaryConfidential& Proprietary UPDATEDEPLOYEVALUATETUNE ML MODEL PARAMETERS ML MODEL DESIGN DATA PREPROCESSING State of the Industry: Complex & Time Intensive Large computational resource . Machine learning expertise . Manual data labeling
  • 10.
    Google is anAI company
  • 11.
    Confidential & ProprietaryGoogleCloud Platform 11 Rapidly accelerating use of deep learning at Google Google Cloud Platform 11 Google directories containing Brain Models 2012 2013 2014 2015 3000 2000 1000 0 Used across products: 4000 2016 Uniqueprojectdirectories 2017
  • 12.
    Confidential & Proprietary DemocratizeAI by making it accessible, fast and useful for enterprises and developers
  • 13.
    © 2017 GoogleInc. All rights reserved.© 2018 Google LLC. All rights reserved. Three Flavors of Machine Learning
  • 14.
    Proprietary + Confidential MLFrameworks: Total Control for Power Users Machine Learning APIs: Ready to Go Cloud Vision API Cloud Translation API Cloud Natural Language API Cloud Speech API Cloud ML EngineTensorFlow Cloud Jobs API Cloud Video Intelligence API AutoML: Bring Your Own Data (We Do the Rest) Pick Your Flavor Cloud DataprocSpark ML
  • 15.
    Proprietary + Confidential Easeof Use vs Flexibility
  • 16.
    Proprietary + Confidential MLFrameworks: Total Control for Power Users Machine Learning APIs: Ready to Go Cloud Vision API Cloud Translation API Cloud Natural Language API Cloud Speech API Cloud ML EngineTensorFlow Cloud Jobs API Cloud Video Intelligence API AutoML: Bring Your Own Data (We Do the Rest) Pick Your Flavor Cloud DataprocSpark ML
  • 17.
  • 18.
    Vision API Detect broadsets 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
  • 19.
    Natural Language API Identifyentities 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
  • 20.
    Speech API Powered bydeep 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
  • 21.
    Translation API Supports morethan 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
  • 22.
    Video Intelligence API Detectentities 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
  • 23.
    @glaforge @manekinekko Cross-platform toolfor building advanced conversational interfaces ListSessions( topics.ComputeEngine); “What Compute Engine talks are there?”
  • 24.
    @glaforge @manekinekko Integrate with... Actionson Google ● Google Home, Pixel… ● and more to come External integrations ● Slack, Facebook Messenger, ● Twitter, Twilio, Skype, Tropo, ● Telegram, Kik, LINE, Cisco Spark, ● Alexa, Cortana
  • 25.
  • 26.
  • 27.
    Proprietary + Confidential MLFrameworks: Total Control for Power Users Machine Learning APIs: Ready to Go Cloud Vision API Cloud Translation API Cloud Natural Language API Cloud Speech API Cloud ML EngineTensorFlow Cloud Jobs API Cloud Video Intelligence API AutoML: Bring Your Own Data (We Do the Rest) Pick Your Flavor Cloud DataprocSpark ML
  • 28.
  • 29.
    Confidential & ProprietaryGoogleCloud Platform 29 AI expertise + Data + Computation We call this AutoML Current solution But we can turn this into Data + 100x Computation
  • 30.
    How does AutoMLwork? Controller: proposes ML models Train & evaluate models 20K times Iterate to find the most accurate model
  • 31.
    Confidential & ProprietaryGoogleCloud Platform 31 AI does AI Systematic exploration of the model space, using the techniques finessed in AlphaGo, yields super-human performance in AI network design
  • 32.
  • 33.
    AutoML for CloudCustomers Dataset Baseline AutoML Customer 1 (Media) 75% 99% Customer 2 (Housing) 87% 93% Customer 3 (Wildlife) 85% 95% Customer 4 (Sports) 90% 96% Customer 5 (Insurance) 0.7 mean AUC 0.95 mean AUC Results for AutoML on image problems https://cloud.google.com/automl/
  • 34.
    Proprietary + Confidential MLFrameworks: Total Control for Power Users Machine Learning APIs: Ready to Go Cloud Vision API Cloud Translation API Cloud Natural Language API Cloud Speech API Cloud ML EngineTensorFlow Cloud Jobs API Cloud Video Intelligence API AutoML: Bring Your Own Data (We Do the Rest) Pick Your Flavor Cloud DataprocSpark ML
  • 35.
  • 36.
    ● PaaS forTensorflow ● Instantly scale your training up to 100 workers (industry leading) ● Automatic monitoring and logging ● Seamlessly transition from training to prediction ● Built in model version management ● No lock-in. Option to download your trained models for on-premise or mobile deployment Cloud ML Engine
  • 37.
    Automatically tune yourmodel with HyperTune ● Automatic hyperparameter tuning service ● Build better performing models faster and save many hours of manual tuning ● Google-developed search algorithm efficiently finds better hyperparameters for your model/dataset ● Flexible: Hyperparameters are provided to to user code as command line flags, allowing any post-processing you want. HyperParam #1 Objective Want to find this Not these HyperParam #2
  • 38.
  • 39.
    Google Use OfTensorFlow: # of Models Search Gmail Translate Maps Android Photos Speech YouTube Play … many others ... Production use in many areas: Research use for: 100s of projects and papers Internal TensorFlow launch
  • 41.
    Google-designed custom ASICbuilt and optimized for TensorFlow 1st generation used in production for over 16 months Now on 2nd generation—180 Teraflops per TPU TensorFlow Research Cloud—1000 TPUs for researchers, at no charge. Tensor Processing Unit
  • 42.
    TPU Pod 64 2nd-genTPUs 11.5 petaflops 4 terabytes of memory 2-D toroidal mesh network
  • 43.
    © 2017 GoogleInc. All rights reserved.© 2018 Google LLC. All rights reserved. Building custom ML solutions on Google Cloud
  • 44.
    Proprietary + Confidential DefineML use cases Define specific ML use cases for the project Select algorithm Choose the right ML algorithm for the task Build ML model Develop the first iteration of the ML model Present results Present results of the model in a way that demonstrates its value to stakeholders Iterate ML model Refine the ML model to improve performance and efficacy Data pipeline & feature engineering Create the right features from raw data for the ML task Plan for deployment Prepare for deployment in production Operationalize model Deploy and operationalize ML model in production Monitor model Monitor deployed ML model and retrain or rebuild when performance degrades 1 3 10 789 Data exploration Perform exploratory data analysis to understand the data 2 4 6 5 Start a new ML project with PSO Cloud Discover Cloud MVM Cloud Deploy Machine Learning Lifecycle
  • 45.
    Building an MLmodel requires 3 things Data Compute Talent Data Scientist Software Engineer
  • 46.
  • 47.
    It's not whohas the best algorithm who wins, it's who has the most data. — Andrew Ng , Co-Founder of Google Brain “ ”
  • 48.
  • 49.
    ● Easily accessand analyze public and commercial datasets hosted on GCP Commercial Datasets Program
  • 50.
  • 51.
  • 52.
    Talent ● Every organizationhas people capable of building ML systems ● But those people may not have the training and tools they need to be successful with machine learning ● Google provides both
  • 53.
    Training ● Google ProfessionalServices will bring Google Machine Learning expertise to your company ● Intensive trainings and workshops from 1 to 4 weeks ● Customized to your needs Talent
  • 54.
    Proprietary + Confidential WeCan Help You Implement your Solution Google PSO1 Google + Partner2 3 Implement Machine Learning APIs Build ML Models Cloud ML Engine Deploy and ManageAnalyze and Plan Google Cloud Google Cloud
  • 55.
    Proprietary + Confidential DefineML use cases Define specific ML use cases for the project Select algorithm Choose the right ML algorithm for the task Build ML model Develop the first iteration of the ML model Present results Present results of the model in a way that demonstrates its value to stakeholders Iterate ML model Refine the ML model to improve performance and efficacy Data pipeline & feature engineering Create the right features from raw data for the ML task Plan for deployment Prepare for deployment in production Operationalize model Deploy and operationalize ML model in production Monitor model Monitor deployed ML model and retrain or rebuild when performance degrades 1 3 10 789 Data exploration Perform exploratory data analysis to understand the data 2 4 6 5 Start a new ML project with PSO Cloud Discover Cloud MVM Cloud Deploy Machine Learning Lifecycle
  • 56.
    © 2018 GoogleLLC. All rights reserved. 2-day workshop, up to 5 days Objectives ● Gain new competitive advantages with ML ● Identify potential ML use cases ● Address targeted business problems ● Survey the foundation for ML potential Activities and Deliverables 1. ML overview session 2. Use case ideation workshop 3. High-level data qualification 4. Analysis and recommendations Cloud Discover: Machine Learning Cloud Discover: Machine Learning helps you understand machine learning (ML) concepts and identify and qualify potential business problems that can be addressed using ML. Ideal for determining if ML is right for your business and which use cases are realistic and achievable.
  • 57.
    © 2018 GoogleLLC. All rights reserved. 2+ months Engagement activities ● Develop an ML solution model ● Perform exploratory data analysis ● Explore the right set of features to include ● Present results of the model that shows its business value Deliverables 1. Data exploration 2. Data pipeline and feature engineering 3. Build and iterate ML model 4. Present results to stakeholders Cloud Deploy ML for MVM Cloud Deploy ML for MVM (minimum viable model) helps you take the use case identified in Cloud Discover from theoretical to practical by developing an ML model. This model will prove the value of the use case and its ML solution to stakeholders prior to investing more in the next phase of build out.
  • 58.
    © 2018 GoogleLLC. All rights reserved. Cloud Deploy ML for MVM details Implements steps 2–7 in the ML lifecycle Delivers a minimum viable ML model or feasibility study Delivers 32 days of work effort in a 2-month calendar window Prices start at $160K for 1 SCE at 100% and 1 consultant at 50%
  • 59.
    © 2018 GoogleLLC. All rights reserved. ● Data exploration ● Algorithm selection ● Data pipeline ● Feature engineering ● Development of initial ML model ● Iteration to improve performance of ML model ● Building a complete data pipeline ● Deploying the model into production ● Converting the model into an API In scope: Out of scope: Cloud Deploy ML for MVM scope
  • 60.
    © 2018 GoogleLLC. All rights reserved. How do we deploy? (detailed level) 32-day work effort (minimum) Advisory on ML model deployment Prerequisite: Discover ML Typical machine learning journey Assess Accelerate Transform Cloud Discover Assessment What do we? Can we? 7-day FTE work effort ML concepts training Group ideation with business and data owners of use cases Use case technical qualification Next steps on how to create a test model for the use cases • • • • • • Step 1 should be taken to determine a feasible ML use case. Qualified customers can proceed directly to Step 3 unless they need training or are not fully committed to starting a formal ML project. Data Engineering /GCP class What is ML? How to use data? 4-day training w/ hands-on lab • • ASL: Immersion Education How to code ML? 1-month immersive training at Google using public datasets • • Cloud Deploy ML for MVM • • • • ASL: Solution Development Let’s build an ML model together 6-month solution development w/Google engineers for one use case Prerequisite: qualified use case • • • Cloud Deploy ML for Production How do we deploy? (detailed level) 20-day work effort (minimum) Advisory on ML model deployment Prerequisite: ML MVM • • • •
  • 61.
  • 62.
    Confidential & Proprietary ProductionRecommendation Solution on GCP Google Analytics BigQuery Google Analytics 360 Customer Web Application Web Server Application Server Database Server Rec API App Engine Cloud Endpoints Model Training Cloud Machine Learning Orchestration Cloud Composer ML Data Training Model files Browser Client Mobile / Tablet Client
  • 63.
    Confidential & Proprietary Kurier.at 3rdLargest news provider in Austria Parent company: 500 web site properties Google Analytics 360 customer Outbrain user… Not happy eDialog GA Partner
  • 64.
    © 2017 GoogleInc. All rights reserved.© 2018 Google LLC. All rights reserved. Customer Success Stories
  • 65.
    Google Cloud Customers:Data & Machine Learning
  • 66.
    © 2018 GoogleLLC. All rights reserved. • Risk analytics and regulation • Fraud detection • Credit worthiness evaluation • Customer segmentation • Cross-selling and upselling • Sales and marketing campaign management Financial Services • Aircraft scheduling • Dynamic pricing • Social media – consumer feedback and interaction analysis • Customer complaint resolution • Traffic patterns and congestion management Travel and Hospitality Explore machine learning use cases with Google • Predictive inventory planning • Recommendation engines • Upsell and cross-channel marketing • Market segmentation and targeting • Customer ROI and lifetime value Retail • Alerts and diagnostics from real-time patient data • Disease identification and risk stratification • Patient triage optimization • Proactive health management • Healthcare provider sentiment analysis Healthcare and Life Sciences • Predictive maintenance or condition monitoring • Warranty reserve estimation • Propensity to buy • Demand forecasting • Process optimization • Telematics Manufacturing • Power usage analytics • Seismic data processing • Carbon emissions and trading • Customer-specific pricing • Smart grid management • Energy demand and supply optimization Energy, Feedstock and Utilities
  • 67.
    complex solve The ability todetect patterns in satellite images — such as the difference between snow and clouds —​is critical to Airbus Defense and Space’s users who depend on highly precise, up-to-date and reliable information. Produits Utilisés Google Cloud Dataflow, BigQuery, Cloud Storage et Cloud Datalab Industry: Aerospace – Region: France “In our tests, Google Cloud Machine Learning enabled us to improve the accuracy and speed at which we analyze the images captured from our satellites. It solved a problem that has existed for decades" . Mathias Ortner, Data Analysis and Image Processing Lead Solutions One of our customers, Airbus Defense and Space, tested the use of Google Cloud Machine Learning to automate the process of detecting and correcting satellite images that contain imperfections such as the presence of cloud formations. problems opportunities identify with advancements such as Machine Learning Speed at which we analyze the images captured Accuracy improved thanks to Machine Learning
  • 68.
    Gain Insights from images bydetecting individual objects and concepts Intelligent structure for 30 billion files managed with powerful capabilities Speed up search and discovery with image-centric workflows Enabled accuracy for automation performance Industry: Technology – Region: North America Box customers unlock new value from content as they automatically classify images and greatly accelerate business processes. Improved Extensive content management for customers in every industry Products Used Google Cloud Vision API Solution Google Cloud Platform and machine learning enable Box to help its customers manage and gain insight from their image files, and speed up image-centric processes and workflows.
  • 69.
    119languages and dialects supported Machinelearning that Enables applications that react to what people say Understand intent of callers in addition to what they say Cutting-edge speech recognition capabilities Products Used Google Cloud Speech API Industry: Technology – Region: North America “Machine Learning has changed the game.” Jeff Lawson, CEO, Twilio Expanded support in Twilio Understand with Google Cloud Speech API Solution Google Speech API enables Twilio to make it easier for developers to build applications that react to what people say during phone calls, taking callers speech and turn it to text. By adding a layer of machine intelligence over existing support, customers can bypass navigating menus and using phone keypads.
  • 70.
    Machine learning isa core, transformative way by which we’re rethinking how we’re doing everything. - Sundar Pichai “ ”
  • 71.