AWS offers a family of intelligent services that provide cloud-native machine learning and deep learning technologies to address different use cases and needs. This deck will help you to gain insight into practical use cases for Amazon Lex, Amazon Polly, and Amazon Rekognition, and learn about newly announced services Amazon Rekognition Video, Amazon Comprehend, Amazon Translate, and Amazon Transcribe. This presentation took place in Australia and New Zealand as part of the AWS Learning Series in 2018.

1. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Craig Stires, Head of AI, Analytics, Big Data - Asia Pacific, AWS
Olivier Klein, Head of Emerging Tech - Asia Pacific, AWS
March 2018
Unlocking New Todays
AI and ML on AWS

2. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
What was possible ... changed

3. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Machine Learning accessible to many more
people
A few highly specialized individuals
Not long ago
Every developer and data scientist
Today

4. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
AW S DEEP LEARNING AMI
Apache
MXNet
TensorFlowCaffe2 Torch KerasCNTK PyTorch GluonTheano
VISION LANGUAGE
Amazon Rekognition
Image
Amazon
Comprehend
Amazon Lex
Amazon Rekognition
Video
AWS DeepLensAmazon SageMaker Spark and EMR
Amazon Polly
Amazon
Translate
Amazon
Transcribe
The Amazon machine learning stack
PLATFORM SERVICES
FRAMEWORKS AND INTERFACES
APPLICATION SERVICES

5. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
3 ways our customers are innovating

6. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Seamless experience - no more boundaries
Seamless
experience
Remove the hard edges that start and stop an experience
Provide customers with an immersive, natural experience that
starts and stops without artificial boundaries. These
experiences include voice enablement, movement detection,
intelligent visual context, and are responsive to intention.
Example Use Cases
Chatbots and roboadvisors Service fulfilment via Alexa Skills
Sentiment detection and response Immersive, enriched multimedia
Checkout-free shopping AI-enhanced sales associates
Facial detection for personalization Fraud detection and prevention

7. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Autonomous machines - reliability and safety
Mechanized labor increases the ability to scale safely
Transform operations through autonomous systems -- offloading to
machines the undifferentiated or unsafe human tasks and
decision making. Systems can perform more efficiently,
accurately, and safely at scale.
Autonomous
Machines
Example Use Cases
Predictive maintenance Autonomous vehicles
Shop floor automation Zone access control and event ID
Robotic emergency response Ambient change detection

8. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Scientific breakthrough - golden age of discovery
Solve fundamental needs for humans, animals, and the earth
Scientific discoveries using deep learning require tremendous
compute capabilities to build and validate models -- now available.
Frameworks and development platforms allow for a broadened
community of users from scientists to analysts.
Scientific
Breakthrough
Example Use Cases
Patient treatment & safety Agricultural yield optimization
Urban maintenance Livestock health and production
Quantum / particle advancements Utility grid efficiency

9. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
C U S T O M E R S R U N N I N G M A C H I N E L E A R N I N G O N A W S T O D A Y

10. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Building AI/ML into business solutions

11. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
AW S DEEP LEARNING AMI
Apache
MXNet
TensorFlowCaffe2 Torch KerasCNTK PyTorch GluonTheano
VISION LANGUAGE
Amazon Rekognition
Image
Amazon
Comprehend
Amazon Lex
Amazon Rekognition
Video
AWS DeepLensAmazon SageMaker Spark and EMR
Amazon Polly
Amazon
Translate
Amazon
Transcribe
The Amazon machine learning stack

12. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Ingest ServingData
sources
Modern data architecture
Insights to enhance business applications, new digital services
Transactions
Web logs /
cookies
ERP
Data analysts
Data scientists
Business users
Engagement platformsConnected
devices
Automation / events
DATA PIPELINES
EVENT PIPELINES
Data
Event
Insights
Data
Lake
Social media

13. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Ingest ServingData
sources
Modern data architecture
Insights to enhance business applications, new digital services
Transactions
Web logs /
cookies
ERP
Data analysts
Data scientists
Business users
Engagement platformsConnected
devices
Automation / events
DATA PIPELINES
EVENT PIPELINES
Data
Event
Insights
Data
Lake
ML / Analytics
Social media

14. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Ingest ServingData
sources
Modern data architecture
Insights to enhance business applications, new digital services
Transactions
Web logs /
cookies
ERP
Data analysts
Data scientists
Business users
Engagement platformsConnected
devices
Automation / events
DATA PIPELINES
EVENT PIPELINES
Data
Event Action
Insights
Data
Lake
ML / Analytics
Predict /
Recommend
AI Services
Social media

15. Building with Artificial Intelligence

16. Hardware &
Learning
algorithms
Fit a function
to the data
Fit a network
structure to
the data

17. Statistical
Machine Learning
Deep Learning
Machine Learning
Feature
engineeringData
Data
Automatically
identifies features
Model
Model
Unseen Sample
Prediction
Unseen Sample
Prediction

18. Artificially Intelligent
+
+
Unique Individuality
Interact Naturally

19. Natural Language
Understanding Computer VisionVoice and ASR
Forecasting Recommendations
Interact Naturally
Unique Individuality
Clustering

20. Alexa, Hello!

21. Extend capabilities with the help of an
open ecosystem!
Amazon
Echo
Alexa Skill AWS Lambda External API Service

22. AW S DEEP LEARNING AMI
Apache
MXNet
TensorFlowCaffe2 Torch KerasCNTK PyTorch GluonTheano
VISION LANGUAGE
Amazon Rekognition
Image
Amazon
Comprehend
Amazon Lex
Amazon Rekognition
Video
AWS DeepLensAmazon SageMaker Spark and EMR
Amazon Polly
Amazon
Translate
Amazon
Transcribe
The Amazon machine learning stack

23. Amazon Transcribe: Automatic Speech Recognition
Time stamps and
confidence scores
Support for both
regular and
telephony audio
Punctuation
§
Detect multiple
speakers
Custom
vocabulary

24. Amazon Comprehend: NLP
😃
Sentiment Entities Languages Key phrases Topic modeling
POWERED BY
DEEP LEARNING

25. Amazon.com, Inc. is located in Seattle, WA
and was founded July 5th, 1994 by Jeff
Bezos. Our customers love buying
everything from books to blenders at great
prices
Named Entities
• Amazon.com: Organization
• Seattle, WA : Location
• July 5th,1994: Date
• Jeff Bezos : Person
Keyphrases
• Our customers
• books
• blenders
• great prices
Sentiment
• Positive
Language
• English
Text Analysis

26. Call center insights
Transcribe 8Khz call
recordings with
high accuracy
Analyze the text with
Amazon Comprehend
Visualize results
on Amazon
QuickSight
Amazon
Transcribe
Amazon
Comprehend
Amazon
Connect
Amazon
Quicksight

27. Amazon Lex
Conversational interfaces for your
applications, powered by the same
Natural Language Understanding
(NLU) & Automatic Speech
Recognition (ASR) models as Alexa
Voice and text
“chatbots”
Integrates with
call centers
Voice interactions
on mobile, web,
and devices
Text interaction
with Slack, Twilio SMS, Kik,
Facebook Messenger
Enterprise
connectors

28. Intents
A particular goal that
the user wants to
achieve
Utterances
Spoken or typed phrases
that invoke your intent
Slots
Data the user must provide to fulfill the
intent
Prompts
Questions that ask the user to
input data
Fulfillment
The business logic required to fulfill the
user’s intent
BookHotel

29. Amazon Rekognition
Object and Scene
Detection
Facial
Analysis
Face
Recognition
Text in Image
Deep learning-based computer vision service for images and video
Unsafe Image
Detection
Celebrity
Recognition

30. Amazon
Rekognition
Videos
Pictures
Objects
Scenes
Sentiments
Emotions
Celebrities
Faces

32. Station DepartureArrival

33. Station DepartureArrival

34. Arrival Station Departure

35. Computer Vision Implementation to identify
empty space and detect object misplacements

36. A fully managed service that enables data scientists and
developers to quickly and easily build machine-learning based
models into production smart applications.
Amazon SageMaker
End-to-End Machine
Learning Platform
In-Built ML
Algorithms
Flexible Model
Training
Pay by the second

37. Amazon SageMaker
BUILD
DEPLOY TRAIN

38. Deploy model
in production
Train and tune
models
Setup &
Manage
environment
for training
Choose &
optimise your
ML model
Scale and
manage the
production
environment
Amazon SageMaker
Pre-built
notebooks for
common
problems
BUILD

39. Amazon SageMaker
One-click
Training
TRAIN
Deploy model
in production
Train and tune
models
Scale and
manage the
production
environment
Built-in, High
performance
Algorithms
Pre-built
notebooks for
common
problems
BUILD

40. SageMaker In-Built Algorithms
K-means Clustering
PCA
Neural Topic Modelling
Factorisation Machines
Linear Learner – Regression
XGBoost
Latent Dirichlet Allocation
Image Classification
Seq2Seq
Linear Learner
Binary Classification
DeepAR Forecasting

41. K-Means Clustering

42. Matrix Factorization

43. Retrain
classification
Feature Extraction
Image Classification Transfer Learning
Convolutional Neural Network

44. MXNet & TensorFlow
SDK
TensorFlow SDK
MXNet (Gluon) SDK
SageMaker Built-in
Algorithms
K-means Clustering
PCA
Neural Topic Modelling
Factorisation Machines
Linear Learner – Regression
XGBoost
Latent Dirichlet Allocation
Image Classification
Seq2Seq
Linear Learner –
Classification
DeepAR Forecasting
Bring Your
Own
Algorithms
ML Algorithms
R
MXNet
TensorFlow
Caffe
PyTorch
Keras
CNTK
…
Apache Spark
Estimator
Apache Spark Python library
Apache Spark Scala library
Training Using Amazon SageMaker
Amazon
EMR

45. Amazon SageMaker
TRAIN
Deploy model
in production
Scale and
manage the
production
environment
Pre-built
notebooks for
common
problems
BUILD
Built-in, High
performance
Algorithms
One-click
Training
Hyperparameter
Optimisation

46. Amazon SageMaker
One-click
Deployment
Scale and
manage the
production
environment
DEPLOY
Pre-built
notebooks for
common
problems
Built-in, High
performance
Algorithms
One-click
Training
Hyperparameter
Optimisation
BUILD TRAIN

47. Train model in the cloud Run model at the edge
AWS
Greengrass
AWS IoT
Tesla V100
120 TFLOPS
Amazon
EC2

48. Amazon SageMaker
Fully managed
hosting with
Auto-scaling
DEPLOY
One-click
Deployment
One-click
Deployment
Pre-built
notebooks for
common
problems
Built-in, High
performance
Algorithms
One-click
Training
Hyperparameter
Optimisation
BUILD TRAIN

49. Amazon
Sagemaker
AWS
Lambda
Greengrass
Core
Alexa Skill
Shadow
Edge
Amazon Echo
object-detection

50. Join us to learn how Amazon and pioneering
businesses build a culture which drives
innovation at scale. Register Today!
#AWSOnAir
@AWSCloudANZ
#AWSSummit