3. A Flywheel For Data
Better Products
More Data Better Analytics
4. A Flywheel For Data
More Users Better Products
More Data Better Analytics
5. A Flywheel For Data
Click stream
User activity
Generated content
Purchases
Clicks
Likes
Sensor data
More Users Better Products
More Data Better Analytics
6. A Flywheel For Data
Click stream
User activity
Generated content
Purchases
Clicks
Likes
Sensor data
Object Storage
Databases
Data warehouse
Streaming analytics
BI
Hadoop
Spark/Presto
Elasticsearch
More Users Better Products
More Data Better Analytics
7. A Flywheel For Data
Click stream
User activity
Generated content
Purchases
Clicks
Likes
Sensor data
Object Storage
Databases
Data warehouse
Streaming analytics
BI
Hadoop
Spark/Presto
Elasticsearch
Artificial
Intelligence
More Users Better Products
More Data Better Analytics
9. Artificial Intelligence At Amazon
Thousands Of Employees Across The Company Focused on AI
Discovery &
Search
Fulfilment &
Logistics
Enhance
Existing Products
Define New
Categories Of
Products
Bring Machine
Learning To All
16. The Advent Of
Deep Learning
Data
GPUs
& Acceleration
Programming
models
Algorithms
17. A Stack for ML/DL Applications
Managed API Services
ML Platform (Data Science Environment) - Notebooks, Model
Hosting and Retraining
DL Engine – MXNet, NeMo, TesorFlow, Caffe, Torch, Theano
Hardware – Distributed Computing, GPU, FPGA
CONTROL
USABILITY&SIMPLICITY
18. EC2 P2 Instance | Up to 16 GPUs
Up to 8 NVIDIA Tesla K80 Accelerators, each running a pair of
NVIDIA GK210 GPUs
Each GPU provides 12 GiB of memory (accessible via 240 GB/second
of memory bandwidth), and 2,496 parallel processing cores.
Currently available in PDX, IAD, DUB, and GovCloud Regions
Instance
Name
GPU Count vCPU Count Memory
Parallel
Processing
Cores
GPU Memory
Network
Performance
p2.xlarge 1 4 61 GiB 2,496 12 GiB High
p2.8xlarge 8 32 488 GiB 19,968 96 GiB 10 Gigabit
p2.16xlarge 16 64 732 GiB 39,936 192 GiB 20 Gigabit
19. One-Click GPU
Deep Learning
AWS Deep Learning AMI
Up to~40k CUDA cores
MXNet
TensorFlow
Theano
Caffe
Torch
Pre-configured CUDA drivers
Anaconda, Python3
+ CloudFormation template
+ Container Image
21. Deep Learning Applications
“deep learning” trend in the past 10 years
image understanding speech recognition natural language
processing
…
autonomy
22. DL Applications on AWS
Realtime detection and tracking on TX1
~10 frame/sec with 640x480 resolution
BlindTool by Joseph Paul Cohen
demo on Nexus 4
23. Can We Help Customers
Put Intelligence At The Heart Of
Every Application & Business?
25. Amazon AI: Three New Deep Learning Services
Polly
Life-like Speech
26. Amazon AI: Three New Deep Learning Services
Rekognition
Life-like Speech Image Analysis
Polly
27. Amazon AI: Three New Deep Learning Services
Rekognition Lex
Life-like Speech Image Analysis Conversational
Engine
Polly
28. Amazon AI: Three New Deep Learning Services
Polly Rekognition Lex
Life-like Speech Image Analysis Conversational
Engine
29. The Advent Of Conversational Interactions
1st Gen: Machine-oriented
interactions
30. The Advent Of Conversational Interactions
1st Gen: Machine-oriented
interactions
2nd Gen: Control-oriented
& translated
31. The Advent Of Conversational Interactions
1st Gen: Machine-oriented
interactions
2nd Gen: Control-oriented
& translated
3rd Gen:
Intent-oriented
32. Lex: Build Natural, Conversational Interactions In Voice & Text
Voice & Text
“Chatbots”
Powers
Alexa
Voice interactions
on mobile, web
& devices
Text interaction
with Slack & Messenger
Enterprise
Connectors
(with more coming) Salesforce
Microsoft Dynamics
Marketo
Zendesk
Quickbooks
Hubspot
37. Origin
Destination London Heathrow
Departure Date
Flight Booking
“Book a flight to
London”
Automatic
Speech Recognition
Natural Language
Understanding
Book Flight
London
Utterances
Flight booking
London Heathrow
Intent /
Slot model
38. Origin Seattle
Destination London Heathrow
Departure Date
Flight Booking
“Book a flight to
London”
Automatic
Speech Recognition
Natural Language
Understanding
Book Flight
London
Utterances
Flight booking
London Heathrow
LocationLocation
Intent /
Slot model
39. Origin Seattle
Destination London Heathrow
Departure Date
Flight Booking
“Book a flight to
London”
Automatic
Speech Recognition
Natural Language
Understanding
Book Flight
London
Utterances
Flight booking
London Heathrow
Prompt
LocationLocation
“When would you like to fly?”
Intent /
Slot model
40. Origin Seattle
Destination London Heathrow
Departure Date
Flight Booking
“Book a flight to
London”
Automatic
Speech Recognition
Natural Language
Understanding
Book Flight
London
Utterances
Flight booking
London Heathrow
Prompt
LocationLocation
“When would you like to fly?”
“When would you like to
fly?”
Polly
Intent /
Slot model
43. Origin Seattle
Destination London Heathrow
Departure Date 11/18/2016
Flight Booking
“Next Friday”
Automatic
Speech Recognition
Natural Language
Understanding
Next Friday
Utterances
Flight booking
11/18/2016
Intent /
Slot model
44. Origin Seattle
Destination London Heathrow
Departure Date 11/18/2016
Flight Booking
“Next Friday”
Automatic
Speech Recognition
Natural Language
Understanding
Next Friday
Utterances
Flight booking
11/18/2016
Intent /
Slot model
45. Origin Seattle
Destination London Heathrow
Departure Date 11/18/2016
Flight Booking
“Next Friday”
Automatic
Speech Recognition
Natural Language
Understanding
Next Friday
Utterances
Flight booking
11/18/2016
Confirmation
“Your flight is booked for next Friday”
Intent /
Slot model
46. Origin Seattle
Destination London Heathrow
Departure Date 11/18/2016
Flight Booking
“Next Friday”
Automatic
Speech Recognition
Natural Language
Understanding
Next Friday
Intent /
Slot model
Utterances
Flight booking
11/18/2016
“Your flight is booked for
next Friday”
Confirmation
“Your flight is booked for next Friday”
Polly
47. Origin Seattle
Destination London Heathrow
Departure Date 11/18/2016
Flight Booking
“Next Friday”
Automatic
Speech Recognition
Natural Language
Understanding
Next Friday
Grammar
Graph
Utterances
Flight booking
11/18/2016
Hotel Booking
52. “Today in Seattle, WA, it’s 11°F”
‘"We live for the music" live from the Madison Square Garden.’
1. Automatic, Accurate Text Processing
Polly: A Focus On Voice Quality & Pronunciation
53. Polly: A Focus On Voice Quality & Pronunciation
2. Intelligible and Easy to Understand
1. Automatic, Accurate Text Processing
54. 2. Intelligible and Easy to Understand
3. Add Semantic Meaning to Text
“Richard’s number is 2122341237“
“Richard’s number is 2122341237“
Telephone Number
Polly: A Focus On Voice Quality & Pronunciation
1. Automatic, Accurate Text Processing
55. 2. Intelligible and Easy to Understand
3. Add Semantic Meaning to Text
4. Customized Pronunciation
“My last name is Nguyen.”
Polly: A Focus On Voice Quality & Pronunciation
1. Automatic, Accurate Text Processing
“My last name is Nguyen.”
56. Polly: Life-like Speech Service
High quality,
through
best-in-class
deep learning
Deep
functionality
Easy to use
& thoughtfully integrated
Built for
production
Low
cost
57. Amazon AI: Three New Deep Learning Services
Polly Rekognition Lex
Life-like Speech Image Analysis Conversational
Engine
58. Rekognition: Search & Understand Visual Content
Real-time &
batch image
analysis
Object & Scene
Detection
Face Comparison Face SearchFacial Analysis
59. Object and Scene Detection
Search, filter, and
curate image
libraries
Smart searches for
user generated
content
Photo, travel, real
estate, vacation
rental applications
60. Avoid faces when cropping
images and overlaying ads
Capture user demographics
and sentiment
Recommend the best photos
Improve online dating match
recommendations
Dynamic, personalized ads
Facial Analysis
61. Face Comparison
Add face verification to
applications and devices
Extend physical security
controls
Provide guest access to
VIP-only facilities
Verify users for online
exams and polls
62. Facial Recognition
Add friend tagging to social
and messaging apps
Assist public safety officers
find missing persons
Identify employees as they
access sensitive locations
Identify celebrities in
historical image archives
63. Early Feedback
“it is all very exciting and it was
shocking how easy it was to set up”
“Rekognition simplifies the process of tagging
and organizing end users’ photos to help them
quickly find relevant images.” - RealNetworks