Re:Invent 2019 Recap. AWS User Group Zaragoza. Javier Ramirez

Javier Ramirez (@supercoco9)
AWS Developer Advocate
AWS re:Invent 2019 Launch Announcement Highlights

© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.
https://www.youtube.com/watch?v=2vdg
_45HsTo

Selected
See details and full list of announcements at
https://aws.amazon.com/new/reinvent/

https://aws.amazon.com/about-aws/global-infrastructure
https://aws.amazon.com/cloudfront/features/?p=ugi&l=emea
https://aws.amazon.com/compliance/data-center/data-centers/
https://www.infrastructure.aws/

AWSGraviton2 Processor
Introducing
Enabling the best price/performance for your cloud workloads
Graviton1 Processor
Graviton2
Processor
DRAFTCompute
Preview – December 3
L E A R N M O R E CMP322-R: Deep dive on EC2 instances powered by AWS Graviton Wednesday 9:15am, MGM

Amazon Confidential
AWSGraviton2 Based Instances
Introducing
Up to 40% better price-performance for general purpose, compute intensive, and
memory intensive workloads.
l
M6g C6g R6g
DRAFT
Built for: General-purpose
workloads such as application
servers, mid-size data stores,
and microservices
Instance storage option: M6gd
Built for: Compute intensive
applications such as HPC, video
encoding, gaming, and
simulation workloads
Instance storage option: C6gd
Built for: Memory intensive
workloads such as open-source
databases, or in-memory
caches
Instance storage option: R6gd
Compute
L E A R N M O R E CMP322-R: Deep dive on EC2 instances powered by AWS Graviton Wednesday 9:15am, MGM

Amazon Confidential
SPECcpu2017
• Industry standard CPU
intensive benchmark
• Run on all vCPUs concurrently
• Comparing performance/vCPU
*All SPEC scores estimates, compiled with GCC9 -O3 -march=native,
run on largest single socket size for each instance type tested.
40%
60%
80%
100%
120%
140%
160%
SPECint2017 Rate SPECfp2017 rate
Performance/vCPU
SPECcpu2017 Rate*
M5 M6G
DRAFTCompute

Amazon Confidential
SPECjvm2008
• Java VM benchmark
• Run across all vCPUs concurrently
• Comparing performance/vCPU
*All SPEC scores estimates, run with OpenJDK11 and skipping compiler* and startup.* tests
Tests run on largest single-socket instance size for each instance type tested.
40%
60%
80%
100%
120%
140%
160%
Performance/vCPU
SpecJVM*
M5 M6G
DRAFTCompute

Amazon Confidential
Load Balancing withNginx
40%
60%
80%
100%
120%
140%
160%
Performance(Requests/s)
M5 M6G
Load
Balancer
(nginx)
NGINX v1.15.9, 512 clients, 128 GET/POST payloads, all HTTPS connections,AES128-GCM-SHA256,OpenSSL
1.1.1, 4 target machines, all tests run on 4xl size; load generator c5.9xl; web servers c5.4xls;
All servers run in a cluster placement group
DRAFTCompute

Amazon Confidential
Memcached
40%
60%
80%
100%
120%
140%
160%
Performance(Requests/s)
Throughput
M5 M6G
Memcached
Memcached v1.5.16, 16B keys, 128B values, 7.8M KV-pairs, 576 connections for load generation from 2x c5.9xlarge instances, 16
additional connections used to measure latency from 1 additional c5.9xlarge,;each connection maintains 4096 outstanding
requests; All servers in a cluster placement group
DRAFTCompute

Amazon Confidential
MediaEncoding withx264
• Huge amount of video created daily
• Encoding it reduces bandwidth to
deliver and storage of that video
• Using libx264 encoder encoded
uncompressed 1080p to h264 40%
60%
80%
100%
120%
140%
160%
Performance(Frames/s)
M5 M6G
DRAFTCompute

Amazon Confidential
MachineLearning
• BERT: The state-of-the-art text encoder
• Feature representation (encoding) is crucial for ML
• Deep Neural network for text feature representation
• Graviton2 has fp16 and int8 support to
accelerate Machine Learning workloads
• M6g can outperform M5
• M5 with AVX-512 is limited to FP32
• With FP16 support M6g performs better for CPU
based inference
40%
60%
80%
100%
120%
140%
160%
Improvement(InferenceLatency)
M5 M6G
DRAFTCompute

Amazon Braket
Introducing
Fully managed service that makes it easy for scientists and developers to
explore and experiment with quantum computing.
DRAFTQuantumTechnology
LEARN MORE CMP213: Introducing Quantum Computing with AWS Wednesday 11:30am,Venetian

AWSComputeOptimizer
Introducing
Identify optimal Amazon EC2 instances and EC2 Auto Scaling group for
your workloads using a ML-powered recommendation engine
DRAFTManagementTools
General Availability – December 3
LEARN MORE CMP323:Optimize Performance and Cost forYourAWS Compute Wednesday, 10:45am, MGM

Receive lower rates
automatically. Easy to use
with recommendations in
AWS Cost Explorer
Significant
savings of up to 72%
Flexible across instance family,
size, OS, tenancy or AWS
Region; also applies to AWS
Fargate & soon to AWS
Lambda usage
Compute/Cost Management
LEARN MORE CMP210: Dive deep on Savings Plans Wednesday, 5:30pm
Announced – November 6
Simplify purchasing with a flexible pricing model that offers savings of
up to 72% on Amazon ECS, AWS Fargate & AWS Lambda usage
Savings Plans

DRAFTContainers
LEARN MORE CON-326R - Running Kubernetes Applications on AWS Fargate
Wednesday, 4pm, Aria
Thursday, 1:45pm, MGM
Introducing
The only way to run serverless Kubernetes containers securely,
reliably, and at scale
Amazon EKS for AWS Fargate

Spare capacity with savings
up to 70% off of Fargate
standard pricing
Improved scalability,
reduced operational cost to
run containers
Containers
New Features
Accelerating momentum for AWS container services

Build and maintain secure OS images more quickly & easily
Introducing
DRAFTCompute
EC2 Image Builder

AWS LicenseManager -SimplifiedWindows &
SQLServer BYOL
New Feature
DRAFTCompute
• Bring your eligibleWindows and SQL BYOL Licenses
to AWS
• Leverage existing licensing investments to save costs
• Automate ongoing management of EC2 Dedicated
Hosts
LEARN MORE
WIN201 - Leadership session: Five New Features of Microsoft and .NET on
AWS that you want to learn
Tuesday, 4pm, MGM

Introducing
DRAFTCompute
Helps customers upgrade
legacy applications to run on
newer, supported versions of
Windows Server without any
code changes
Future-proof Reduced risk Cost-effective
Improved security
posture on supported,
new OS
Isolate old runtimes
Compliance with industry
regulations
No application
refactoring or recoding
cost
No extended support
costs
Decouple from
underlying OS
Low risk of failure on
subsequent OS updates
Supports all OS version Reduced operating costs
AWS End of support Migration Program forWindows
Server

15 new capabilities on data availability, security, protection, and cost
optimization across the storage portfolio
DRAFTStorage
Announcement – November 20
SERVICE ANNOUNCEMENTS
Amazon S3 Replication Time Control
Amazon EBS Fast Snapshot Restore (FSR)
Amazon FSx for
File Server
Multi-AZ deployment option; Data Deduplication; User Quotas; Programmatic Share
Management; Support for smaller file systems; Support for High Availability SQL
deployments; Fine-grained control of in-transit encryption
AWS Storage Gateway High Availability for VMware; Performance increase for Tape and File Gateway
AWS DataSync 68% price reduction; Task Scheduling; Expansion to 5 additional AWS regions
Amazon EFS Now available in all commercial AWS regions except China regions

AmazonS3Access Points
Introducing
Simplify managing data access at scale for applications using shared data sets
on Amazon S3. Easily create hundreds of access points per bucket, each with a
unique name and permissions customized for each application.
DRAFTStorage

Amazon ManagedApacheCassandraService
Introducing
A scalable, highly available, and serverless Apache Cassandra–compatible database
service. Run your Cassandra workloads in the AWS cloud using the same Cassandra
application code and developer tools that you use today.
Apache Cassandra-
compatible
Performance
at scale
Highly available
and secure
No servers
to manage
DRAFTDatabases
LEARN MORE DAT324: Overview of Amazon ManagedApache Cassandra Service

DRAFTDatabases
Amazon Aurora Machine Learning Integration
Simple, optimized, and secure Aurora, SageMaker, and Comprehend (in
preview) integration. Add ML-based predictions to databases and applications
using SQL, without custom integrations, moving data around, or ML
experience.

Amazon RDS Proxy
Introducing
Fully managed, highly available database proxy feature for Amazon RDS.
Pools and shares connections to make applications more scalable, more
resilient to database failures, and more secure.
DRAFTDatabases
Public Beta – December 3
LEARN MORE DAT368: Setting up database proxy servers with RDS Proxy

UltraWarm forAmazon ElasticsearchService
Introducing
A low cost, scalable warm storage tier for Amazon Elasticsearch Service. Store up to
10 PB of data in a single cluster at 1/10th the cost of existing storage tiers, while still
providing an interactive experience for analyzing logs.
DRAFTAnalytics
LEARN MORE ANT229:Scalable, secure, and cost-effective log analytics

DRAFTAnalytics
Amazon Redshift RA3 instances with Managed Storage
Optimize your data warehouse costs by paying for compute and storage separately
L E A R N M O R E
ANT213-R1: State of the Art Cloud Data Warehousing
ANT230: Amazon Redshift Reimagined: RA3 and AQUA
Wednesday, 10am, Venetian
Delivers 3x the performance of existing cloud DWs
2x performance and 2x storage as similarly priced
DS2 instances (on-demand)
Automatically scales your DW storage capacity
Supports workloads up to 8PB (compressed)
COMPUTE NODE
(RA3/i3en)
SSD Cache
S 3 S T O R A G E
COMPUTE NODE
(RA3/i3en)
SSD Cache
COMPUTE NODE
(RA3/i3en)
SSD Cache
COMPUTE NODE
(RA3/i3en)
SSD Cache
Managed storage
$/node/hour
$/TB/month
Introducing

AQUA(AdvancedQueryAccelerator)forAmazonRedshift
Introducing
Redshift runs 10x faster than any other cloud data warehouse without increasing cost
DRAFTAnalytics
Private Beta – December 3
LEARN MORE ANT230:Amazon Redshift Reimagined: RA3 and AQUA Wednesday, 10am,Venetian
AQUA brings compute to storage so data doesn't have to
move back and forth
High-speed cache on top of S3 scales out to process data in
parallel across many nodes
AWS designed processors accelerate data compression,
encryption, and data processing
100% compatible with the current version of Redshift
S 3
S T O R A G E
AQUA
ADVANCED QUERY ACCELERATOR
R A 3 C O M P U T E C L U S T E R

Amazon Redshift Data Lake Export
New Feature
No other data warehouse makes it as easy to gain new insights from all
your data.
DRAFTAnalytics
LEARN MORE
ANT335R: How to build your data analytics stack at scale withAmazon
Redshift
Monday, 7pm,Venetian
Tuesday, 11:30am,Aria

Amazon Redshift FederatedQuery
Analyze data across data warehouse, data lakes, and operational
database
New Feature
DRAFTAnalytics
LEARN MORE ANT213-R1: State of the Art Cloud DataWarehousing Tuesday, 3pm, Bellagio

AmazonAthenaFederatedQuery
RunSQLqueriesondataspanningmultipledatastores
Redshift
Data warehousing
ElastiCache
Redis
Aurora
MySQL, PostgreSQL
DynamoDB
Key value, Document
DocumentDB
Document
S3/Glacier
Run connectors in AWS Lambda: no servers to manage
Run SQL queries on relational, non-relational, object,
or custom data sources; in the cloud or on-premises
Open Source Connectors for common data sources
Build connectors to custom data sources
PREVIEW
NEW

AWS Nitro Enclaves
Introducing
Create additional isolation to further protect highly sensitive data within
EC2 instances
Nitro Hypervisor
Instance A Enclave A Instance B
EC2 Host Additional isolation
within an EC2 instance
Isolation between EC2
instances in the same host
Local socket
connection
DRAFTCompute

DRAFTManagement Tools
Identify unusual activity in your AWS accounts
 Save time sifting through logs
 Get ahead of issues before
they impact your business
CloudTrail Insights
Introducing
• Unexpected spikes in resource
provisioning
• Bursts of IAM management
actions
• Gaps in periodic maintenance
activity
L E A R N M O R E MGT420-R: CloudTrail Insights: Identify and Solve Operational

AWS Detective
Introducing
Quickly analyze, investigate, and identify the root cause of security
findings and suspicious activities.
Automatically distills &
organizes data into a
graph model
Easy to use visualizations
for faster & effective
investigation
Continuously updated as
new telemetry becomes
available
DRAFTSecurity
LEARN MORE SEC312: Introduction to Amazon Detective Thursday, 1:45pm,Venetian

AWS IAMAccessAnalyzer
Introducing
Continuously ensure that policies provide the intended public and cross-account access to
resources, such as Amazon S3 buckets, AWS KMS keys, & AWS Identity and Access
Management roles.
DRAFTSecurity
Uses automated reasoning, a form of
mathematical logic, to determine all possible
access paths allowed by a resource policy
Analyzes new or updated resource
policies to help you understand
potential security implications
Analyzes resource policies for
public or cross-account access
LEARN MORE SEC309: Deep Dive into AWS IAM Access Analyzer Thursday, 3:15pm,Venetian

1
Create or use existing identities, including Azure AD, and manage
access centrally to multiple AWS accounts and business applications,
for easy browser, command line, or mobile single sign-on access by
employees.
New Feature
AWS Single Sign-On
DRAFTSecurity
LEARN MORE SEC308: Manage federated user permissions at scale with AWS Thursday, 12:15pm,

Existing Service
DRAFTNetworking
Scale connectivity across
thousands of Amazon VPCs, AWS
accounts, and on-premises
networks
Amazon VPCAmazon VPC
Amazon VPCAmazon VPC
Customer
gateway
VPN
connection
AWS Direct
Connect Gateway
L E A R N M O R E NET203-L Leadership Session Networking Wednesday, 11:30am, MGM
AWS Transit Gateway

New Feature
AWSTransitGateway Inter-Region Peering
DRAFTNetworking
AWS TRANSIT
GATEWAY
Inter-Region Peering
Build global networks by connecting transit gateways across multiple AWS Regions

High availability and improved performance of site-to-site VPN
New Feature
AWS Accelerated Site-to-Site VPN
DRAFTNetworking

AWSTransitGateway Network Manager
Introducing General Availability – December 3
DRAFTNetworking
L E A R N M O R E NET212 -AWSTransit Gateway Network Manager

New Feature
TransitGateway Multicast
DRAFTNetworking
Build and deploy multicast applications in the cloud

New Feature
AmazonVPC Ingress Routing
DRAFTNetworking
Route inbound and outbound traffic through a third party or AWS service

L E A R N M O R E SVS401 - Optimizing your serverless applications
Wednesday, 1:45pm, Mirage
Thursday, 3:15pm,Venetian
ProvisionedConcurrency onAWS Lambda
New Feature
• Keeps functions initialized and hyper-ready, ensuring start
times stay in the milliseconds
• Builders have full control over when provisioned
concurrency is set
• No code changes are required to provision concurrency
on functions in production
• Can be combined with AWS Auto Scaling at launch
DRAFTServerless

Achieve up to 67% cost reduction and 50% latency reduction compared to
REST APIs. HTTP APIs are also easier to configure than REST APIs,
allowing customers to focus more time on building applications.
Reduce application costs by
up to 67%
Reduce application latency
by up to 50%
Configure HTTP APIs easier
and faster than before
HTTP APIs for Amazon API Gateway
Introducing
DRAFTMobile Services
L E A R N M O R E
CON213-L - Leadership session: Using containers and serverless to
modern application development (incl schema registry demo)
Wednesday 9:15am,

AWSStep Functions ExpressWorkflows
Introducing
OrchestrateAWS compute, database, and messaging services at rates
greater than 100,000 events/second, suitable for high-volume event
processing workloads such as IoT data ingestion, streaming data
processing and transformation.
DRAFTApp Integration
L E A R N M O R E API321: Event-ProcessingWorkflows at Scale with AWS Step Functions Wednesday, 3:15pm, MGM

Amazon EventBridgeSchema Registry
Introducing
Store event structure - or schema - in a shared central location, so it’s
faster and easier to find the events you need. Generate code bindings right
in your IDE to represent an event as an object in code.
DRAFTApp Integration
LEARN MORE
CON213-L - Leadership session: Using containers and serverless to accelerate
modern application development (incl schema registry demo)
Wednesday 9:15am,
Venetian

Amplify for iOS &Android
Introducing
Open source libraries and toolchain that enable mobile developers to build
scalable and secure cloud powered serverless applications.
L E A R N M O R E MOB317 - Speed up native mobile development with AWSAmplify Wednesday, 11:30am,Venetian

Amplify DataStore
New Feature
Multi-platform (iOS/Android/React Native/Web) on-device persistent
storage engine that automatically synchronizes data between mobile/web
apps and the cloud using GraphQL.
L E A R N M O R E MOB402: Build data-driven mobile and web apps withAWS AppSync Wednesday, 2:30pm, Mirage

Whatcustomers are doing withAWS IoT
Remotely monitor
patient health &
wellness applications
Manage energy resources
more efficiently
Enhance safety in
the home, the office,
and the factory floor
Transform transportation with
connected and autonomous
vehicles
Track inventory
levels and manage
warehouse operations
Improve the performance and
productivity of industrial
processes
Build smarter products & user
experiences in homes,
buildings, and cities
Grow healthier crops with
greater efficiencies

Nine new features launched across AWS IoT, including eight on IoT
Day, helping customers easily scale their IoT deployments.
DRAFTInternet of Things
Announcements – Nov 25/Dec 1
SERVICE NEW FEATURES
AWS IoT Greengrass
Stream Manager
Container Support
AWS IoT Core
Fleet Provisioning
Configurable Endpoints
Custom Domains for Configurable Endpoints
Custom Authorizer for MQTT Connections
Alexa Voice Service (AVS) Integration
AWS IoT Device Management Secure Tunneling
AWS IoT SiteWise SiteWise Monitor
AWS IoT Day

AlexaVoiceService(AVS)Integration for IoTCore
New Feature
DRAFTInternet ofThings
Quickly and cost effectively go to market with Alexa built-in capabilities on new categories of products such
as light switches, thermostats, and small appliances.
Accelerate time to market with certified
partner development kits that work with
AVS Integration for IoT Core by default.
Lowers the cost of integrating Alexa Voice
up to 50% by reducing the compute and
memory footprint required
Build new categories of Alexa Built-in
products on resource constrained devices
(e.g. ARM ‘M' class microcontrollers with
<1MB embedded RAM).

ContainerSupport forAWS IoTGreengrass
New Feature
DRAFTInternet ofThings
Deploy containers seamlessly to edge devices
Move containers from the cloud to
edge devices using AWS IoT
Greengrass, without rewriting any
code.
Enables both Docker & AWS
Lambda components to operate
seamlessly together at the edge
UseAWS IoTGreengrass Secrets
Manager to manage credentials for
private container registries.

AWSOutposts
Now Available
Fully managed service that extends AWS infrastructure, AWS services, APIs, and tools to virtually any
connected customer site.Truly consistent hybrid experience for applications across on-premises and cloud
environments. Ideal for low latency or local data processing application needs.
Same AWS-designed infrastructure
as in AWS regional data centers
(built on AWS Nitro System)
delivered to customer facilities
Fully managed, monitored, and
operated by AWS
as in AWS Regions
Single pane of management
in the cloud providing the
same APIs and tools as
in AWS Regions
Compute
LEARN MORE
CMP302-R:AWS Outposts: Extend the AWS experience to on-premises
environments
Wednesday at 11:30am,Aria
Thursday at 3:15pm, Mirage
Friday at 10:45am, Mirage

Amazon Confidential
AdditionalAWSServicesSupported LocallyonOutposts

LocalZones
Introducing
Extend the AWS Cloud to more locations and closer to your end-users to
support ultra low latency application use cases. Use familiarAWS services
and tools and pay only for the resources you use.
DRAFTCompute
The first Local Zone to be released will be located in Los Angeles.

AWSWavelength
Introducing
EmbedsAWS compute and storage inside telco providers’ 5G networks.
Enables mobile app developers to deliver applications with single-digit
millisecond latencies. Pay only for the resources you use.
DRAFTCompute
Announcement – December 3

builds and operates the
technologies that underpinAmazon.com
andAWS

The Amazon Builders’ Library
Architecture, software delivery, and operations
By Amazon’s senior technical executives and engineers
Real-world practices with detailed explanations
Content available for free on the website

LaunchAnnouncements covered inthissession
AmazonEC2Inf1Instances
AWSGraviton2Processor
AWSGraviton2BasedInstances
AmazonBraket
AWSNitroEnclaves
AWSComputeOptimizer
SavingsPlans
AmazonEKSforAWSFargate
AWSFargateSpot
AmazonECSCapacityProviders
AmazonECSCLI2.0
ECSClusterAutoscaling
EC2ImageBuilder
AWSLicenseManager-Simplified
Windows&SQLServerBYOL
AWSEndofsupportMigrationProgramfor
WindowsServer
AmazonS3AccessPoints
EBSDirectAPIsforSnapshots
AmazonManagedApacheCassandra
Service
AmazonAuroramachinelearning
integration
AmazonRDSProxy
UltraWarmforAmazonElasticsearch
Service
AmazonRedshiftRA3instanceswith
ManagedStorage
AQUA(AdvancedQueryAccelerator)for
AmazonRedshift
AmazonRedshiftFederatedQuery
AmazonRedshiftDataLakeExport
AWSDataExchange
CloudTrailInsights
AWSDetective
AWSIAMAccessAnalyzer
AWSSingleSign-On
AWSTransitGatewayInter-RegionPeering
AWSAcceleratedSite-to-SiteVPN
AWSTransitGatewayNetworkManager
TransitGatewayMulticast
AmazonVPCIngressRouting
ProvisionedConcurrencyonAWSLambda
HTTPAPIsforAmazonAPIGateway
AWSStepFunctionsExpressWorkflows
AmazonEventBridgeSchemaRegistry
AmplifyforiOS&Android
AmplifyDataStore
AlexaVoiceService(AVS)IntegrationforIoT
Core
ContainerSupportforAWSIoTGreengrass
AWSOutposts
LocalZones
WavelengthZones

AI & Machine Learning Launches
Javier Ramirez
Developer Advocate
AmazonWeb Services
@supercoco9

Pre:Invent highlights
https://aws.amazon.com/about-aws/whats-new/machine-learning
• Amazon Comprehend: 6 new languages
• AmazonTranslate: 22 new languages
• AmazonTranscribe: 15 new languages, alternative transcriptions
• Amazon Lex: SOC compliance, sentiment analysis,
web & mobile integration with Amazon Connect
• Amazon Personalize: batch recommendations
• Amazon Forecast: use any quantile for your predictions
With region expansion across the board!

Speech Recognition
For Healthcare

IntroducingAmazonTranscribe Medical
Easy-to-UseAccurate Affordable

IntroducingAmazon RekognitionCustomLabels
• Import images labeled by Amazon
SageMaker GroundTruth…
• Or label images automatically based on folder structure
• Train a model on fully managed
infrastructure
• Split the data set for training and validation
• See precision, recall, and F1 score at the end of training
• Select your model
• Use it with the usual Rekognition APIs

Customers areforced tochoose
ML only systems are high speed and low cost,
but do not support nuanced decision making
Human only workflows offer nuanced decision
making, but they’re low speed and high cost.
OR

Customers need
+
MachineLearning and humans working together

A2Iletsyoueasilyimplementhumanreviewinmachine
learningworkflowstoimprovetheaccuracy,speed,and
scaleofcomplexdecisions.
IntroducingAmazonAugmentedAI (A2I)

HowAmazonAugmentedAIworks
Client application
sends input data
AWS AI Service or
custom ML model
makes predictions
Results stored to
your S3
1 2
4
Low confidence predictions
sent for human review
3
High-confidence predictions
returned immediately to client
application
5
Amazon Rekognition
Amazon Textract

Human ReviewWorkforces
Amazon MechanicalTurk
An on-demand 24x7 workforce
of over 500,000 independent
contractors worldwide, powered
by Amazon MechanicalTurk
Private
A team of workers that you have
sourced yourself, including your
own employees or contractors
for handling data that needs to
stay within your organization
Vendors
A curated list of third-party
vendors that specialize in
providing data labeling services,
available via deAWS Marketplace

Fraud detection isdifficult
$$$ billions lost to
fraud each year
Online business prone to
fraud attacks
Bad actors often change
tactics
Changing rules = more
human reviews
Dependent on others to
update detection logic

Fraud detection withML isalsodifficult
Top data scientists are
costly & hard to find
One-size-fits-all models
underperform
Often need to
supplement data
Data transformation +
feature engineering
Fraud imbalance =
needle in a haystack

IntroducingAmazon Fraud Detector
A fraud detection service that makes it
easy for businesses to use machine
learning to detect online fraud in real-
time, at scale

Amazon Fraud Detector – KeyFeatures
Pre-built fraud
detection model
templates
Automatic
creation of
custom fraud
detection models
models
Models learn
from past
attempts to
defraud Amazon
Amazon
SageMaker
integration
One interface to
review past
evaluations and
detection logic

Amazon Fraud Detector –Automated Model Building
1 2 4 5
Training
data in S3
63

Speech & Text Analytics
For Contact Centers

Challengesincontactcenters
• Better visibility into quality of customer interactions
• Cost prohibitive
•
• Timely discovery of emerging issues
• Support for live calls
• End user experience

Introducing Contact Lens For Amazon Connect
Theme
detection
Built-in automatic
call transcription
Automated
contact
categorization
Enhanced
Contact Search
Real-time sentiment
dashboard
and alerting
Presents
recurring
issues based
on
Customer
feedback
Identify call types
such as script
compliance,
competitive
mentions,
and cancellations.
Filter calls of
interest based
on words
spoken and
customer
sentiment
View entire call
transcript directly in
Amazon Connect
Quickly identify
when customers
are having a
poor experience
on live calls
Easily use the power of machine learning to improve the quality of your customer experience
without requiring any technical expertise

TypicalApplicationBuildandRunProcess
Write +
Review
Build +
Test
Deploy Measure Improve
1. Code Reviews require expertise in multiple areas such as knowledge
of AWS APIs,Concurrency, etc.
2. Code analyzer tools require high accuracy.
3. Distributed Cloud application are difficult to optimize.
4. Performance engineering expertise is hard to find.

IntroducingAWSCodeGuru
Built-in code reviews
with intelligent
recommendations
Detect and optimize
expensive lines of
code before
production
Easily identify latency
and performance
improvements
production environment
CodeGuru Reviewer CodeGuru Profiler

CodeGuruReviewer:HowItWorks
Input:
Source Code
Feature Extraction Machine Learning
Output:
Recommendations
Customer provides source
code as input
Java
AWS CodeCommit
Github
Extract semantic features /
patterns
ML algorithms identify similar
code for comparison
Customers see recommendations
as Pull Request feedback

CodeGuru Example – Looping vsWaiting
do {
DescribeTableResult describe = ddbClient.describeTable(new DescribeTableRequest().withTableName(tableName));
String status = describe.getTable().getTableStatus();
if (TableStatus.ACTIVE.toString().equals(status)) {
return describe.getTable();
}
if (TableStatus.DELETING.toString().equals(status)) {
throw new ResourceInUseException("Table is " + status + ", and waiting for it to become ACTIVE is not useful.");
}
Thread.sleep(10 * 1000);
elapsedMs = System.currentTimeMillis() - startTimeMs;
} while (elapsedMs / 1000.0 < waitTimeSeconds);
throw new ResourceInUseException("Table did not become ACTIVE after ");
This code appears to be waiting for a resource before it runs. You could use the waiters feature to help improve efficiency.
Consider using TableExists, TableNotExists. For more information, see https://aws.amazon.com/blogs/developer/waiters-in-
the-aws-sdk-for-java/
Recommendation
Code
We should use waiters instead - will help remove a lot of this code.Developer Feedback

CodeGuruProfiler:HowItWorks
Input:
Live application
stack trace
Application profile
sampling
Pattern matching
Output:
Method names,
Recommendations
and searchable
visualizations
Customer application
runs in production
CodeGuru Profiler
continuously captures
application stack trace
information
CodeGuru Profiler detects
performance inefficiencies in the
live application
Customers see
recommendations in their
automated efficiency reports and
visualizations
Amazon Confidential

CodeGuru ProfilerVisualization– Hotspots
Java.util.zip.Deflator.deflateBytes 61%
LCMS.colorEvent 6%JPEGImageReader.readImage 5%

Employeesspend20%oftheirtime
lookingforinformation.
—McKinsey
20%
44%44%ofthetime,theycannotfind
theinformationtheyneedtodo
theirjob.
—IDC

Introducing Kendra
Easy to find what you are
looking for
Fast search, and
quick to set up
Native connectors
(S3, Sharepoint,
file servers,
HTTP, etc.)
Natural language
Queries
NLU and
ML core
Simple API
and console
experiences
Code samples
Incremental
learning through
feedback
Domain
Expertise

Kendra connectors
…and more coming in 2020

Getting startedwithKendra
Step 1
Create an index
An index is the place where
you add your data sources
to make them searchable
in Kendra.
Step 2
Add data sources
Add and sync your data
from S3, Sharepoint, Box
and other data sources, to
your index.
Step 3
Test & deploy
After syncing your data,
visit the Search console
page to test search &
deploy Kendra in your
search application.

Pre:Invent highlights
https://aws.amazon.com/about-aws/whats-new/machine-learning
• Invoke Amazon SageMaker models in Amazon Quicksight
• Invoke Amazon SageMaker models in Amazon Aurora
• Deploy many models on the sameAmazon SageMaker endpoint

Using Kubernetes for ML is hard to
manage and scale
Build and manage services
within Kubernetes cluster for
ML
Make disparate open-source
libraries and frameworks work
together in a secure and
scalable way
Requires time and expertise from
infrastructure, data science, and
development teams
Need an easier way to
use Kubernetes for ML
+
+
=

Fully managed
infrastructure in SageMaker
Introducing Amazon SageMaker Operators for Kubernetes
Kubernetes customers can now train, tune, & deploy models in
Amazon SageMaker

Machine learning is iterative
involving dozens of tools and
hundreds of iterations
Multiple tools needed for
different phases of the
ML workflow
Lack of an integrated
experience
Large number of iterations
Cumbersome, lengthy processes, resulting in
loss of productivity
+
+
=

IntroducingAmazonSageMakerStudio
The first fully integrated development environment (IDE) for machine learning
Organize, track, and compare
thousands of experiments
Easy experiment
management
Share scalable notebooks
without tracking code
dependencies
Collaboration at scale
Get accurate models for with
full visibility & control without
writing code
Automatic model
generation
Automatically debug errors,
monitor models, & maintain
high quality
Higher quality ML
models
Code, build, train, deploy, &
monitor in a unified visual
interface
Increased
productivity

Data science and collaboration
needs to be easy
Setup and manage resources
Collaboration across
multiple data scientists
Different data science
projects have different
resource needs
Managing notebooks and
collaborating across
multiple data scientists is
highly complicated
+
+
=

IntroducingAmazonSageMakerNotebooks
Access your notebooks in
seconds with your corporate
credentials
Fast-start shareable notebooks
Administrators manage
access and permissions
Share your notebooks
as a URL with a single click
Dial up or down
compute resources
Start your notebooks
without spinning up
compute resources

Data Processing and
Model Evaluation involves a lot of
operational overhead
Building and scaling infrastructure
for data processing workloads is
complex
Use of multiple tools or services
implies learning and
implementing new APIs
All steps in the ML workflow need
enhanced security, authentication
and compliance
Need to build and manage tooling
to run large data processing and
model evaluation workloads
+
+
=

IntroducingAmazonSageMakerProcessing
Analytics jobs for data processing and model evaluation
Use SageMaker’s built-in
containers or bring your own
Bring your own script for
feature engineering
Custom processing
Achieve distributed
processing for clusters
Your resources are created,
configured, & terminated
automatically
Leverage SageMaker’s
security & compliance
features

Managing trials and experiments is
cumbersome
Hundreds of experiments
Hundreds of parameters
per experiment
Compare and contrast
Very cumbersome and
error prone
+
+
=

Introducing Amazon SageMaker Experiments
Experiment
tracking at scale
Visualization
for best results
Flexibility with
Python SDK & APIs
Iterate quickly
Track parameters & metrics
across experiments & users
Organize
experiments
Organize by teams, goals, &
hypotheses
Visualize & compare
between experiments
Log custom metrics &
track models using APIs
Iterate & develop high-
quality models
A system to organize, track, and evaluate training experiments

Debugging and profiling
deep learning is painful
Large neural networks
with many layers
Many connections
Additional tooling for analysis
and debug
Extraordinarily difficult
to inspect, debug, and profile
the ‘black box’
+
+
=

Automatic data
analysis
Relevant data
capture
Automatic error
detection
Improved productivity
with alerts
Visual analysis
and debug
Introducing Amazon SageMaker Debugger
Analyze and debug data
with no code changes
Data is automatically
captured for analysis
Errors are automatically
detected based on rules
Take corrective action
based on alerts
Visually analyze & debug
from SageMaker Studio
Analysis & debugging, explainability, and alert generation

Deploying a model is not the end,
you need to continuously monitor
it in production and iterate
Concept drift due to
divergence of data
Model performance can
change due to unknown
factors
Continuous monitoring of model
performance and data involves a
lot of effort and expense
Model monitoring is
cumbersome but critical
+
+
=

Introducing Amazon SageMaker Model Monitor
Automatic data
collection
Continuous
Monitoring
CloudWatch
Integration
Data is automatically
collected from your
endpoints
Automate corrective
actions based on Amazon
CloudWatch alerts
Continuous monitoring of models in production
Visual
Data analysis
Define a monitoring
schedule and detect
changes in quality against a
pre-defined baseline
See monitoring results,
data statistics, and
violation reports in
SageMaker Studio
Flexibility
with rules
Use built-in rules to
detect data drift or write
your own rules for
custom analysis

Successful ML requires
complex, hard to discover
combinations
Largely explorative &
iterative
Requires broad and
complete
knowledge of ML domain
Lack of visibility
Time consuming,
error prone process
even for ML experts
+
+
=
of algorithms, data, parameters

Introducing Amazon SageMaker Autopilot
Quick to start
Provide your data in a
tabular form & specify target
prediction
Automatic
model creation
Get ML models with feature
engineering & automatic model
tuning automatically done
Visibility & control
Get notebooks for your
modelswith source code
Automatic model creation with full visibility & control
Recommendations &
Optimization
Get a leaderboard & continue
to improve your model

Ground
Truth
Algorithms
& Frameworks
Collaborative
notebooks
ExperimentsDistributed
Training &
Debugger
Deployment,
Monitoring, & Hosting
SageMaker AutoPilot
Build, Train, Deploy Machine Learning Models Quickly at Scale
Reinforcement
Learning
Tuning
& Optimization
SageMaker Studio
Marketplace
for ML
Amazon SageMaker

AWS DeepRacer improvements
• AWS DeepRacer Evo
• Stereo camera
• LIDAR sensor
• New racing opportunities
• Create your own races
• Object Detection & Avoidance
• Head-to-head racing

AWS DeepComposer
• The world’s first machine learning-
enabled musical keyboard
• Compose music using Generative
Adversarial Networks (GAN)
• Use a pretrained model, or train
your own

IntroducingAmazon EC2Inferentia
• Fast, low-latency inferencing at a very low cost
• 64 teraOPS on 16-bit floating point (FP16 and BF16) and mixed-precision data.
• 128 teraOPS on 8-bit integer (INT8) data.
• Neuron SDK: https://github.com/aws/aws-neuron-sdk
• Available in Deep Learning AMIs and Deep Learning Containers
• TensorFlow and Apache MXNet, PyTorch coming soon
Instance Name Inferentia Chips vCPUs RAM EBS Bandwidth
inf1.xlarge 1 4 8 GiB Up to 3.5 Gbps
inf1.2xlarge 1 8 16 GiB Up to 3.5 Gbps
inf1.6xlarge 4 24 48 GiB 3.5 Gbps
inf1.24xlarge 16 96 192 GiB 14 Gbps

DeepGraph Library
https://www.dgl.ai
• Python open source library that helps
researchers and scientists quickly build, train,
and evaluate Graph Neural Networks on their
data sets
• Use cases: recommendation, social networks,
life sciences, cybersecurity, etc.
• Available in Deep Learning Containers
• PyTorch andApache MXNet,TensorFlow coming soon
• Available for training on Amazon SageMaker

DeepJavaLibrary
https://www.djl.ai
• Java open source library,
to train and deploy models
• Framework agnostic
• Apache MXNet for now, more will come
• Train your own model, or use a
pretrained one from the model
zoo

Enjoy!
Build cool stuff,
and please send us feedback!

Thank you!
Javier Ramirez
Developer Advocate
Amazon Web Services
@supercoco9

Re:Invent 2019 Recap. AWS User Group Zaragoza. Javier Ramirez

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Re:Invent 2019 Recap. AWS User Group Zaragoza. Javier Ramirez

Similar to Re:Invent 2019 Recap. AWS User Group Zaragoza. Javier Ramirez (20)

More from javier ramirez

More from javier ramirez (20)

Recently uploaded

Recently uploaded (20)

Re:Invent 2019 Recap. AWS User Group Zaragoza. Javier Ramirez