Trending technology in IBM Think

Multicloud, big data, AI/ML infused, containerized framework and APIs, CyberSecurity, blockchain technology, tokenization

1. February 13, Wednesday

852: Track Keynote Put AI to Work for
usiness with IBM Data Science
ebruary 13, Wednesday,

0 AM - 11:10 AM, Moscone South,
xhibit Level, Hall B | Data £ AI Theater C

266: SPSS and Open Source
sbruary 13, Wednesday, 4:30 PM ~ 5:10 PM
scone South | AI Theatre 1

»bruary 13, Wednesday, 5:30 PM - 6:10 PM
jascone South, Room 304

975: Infuse App with AL

sbruary 13, Wednesday, 5:30 PM - 6:10 PM
joscone South | AI Theatre 2

February 14, Thursday

February 14, Thuraay, 9:50 AM 10:10 AM

Moscone South | San Francisco Ballroom 213

February 14, Thursday,9:30 AM - 10:10 AM
Moscone South | AI Theatre 2

February 14, Thursday, 10:30 AM - 11:10 AM
Moscone South, Level 2 | San Francisco Ballroom 216

6966: Operationalize DS with SPSS Modeler
February 14, Thursday, 11:30 AM - 12:10 PM

AMC Metreon | Screen 14
3232: Power of Platform

February 14, Thursday, 1:30 PM - 2:10 PM
Moscone South San Francisco Ballroom 214

February 14, Thursday, 3:30 PM - 4:10 PM
AMC at Metreon, Level 3 | Theatre 14
6962: Watson Studio Desktop

February 14, Thursday, 3:30 PM - 4:10 APM
Moscone South | Room 303

February 14, Thursday, 4:30 PM - 5:10 PM
Moscone South Ballroom 216

ebruary 15, Pnday, 11:30 AM - 12:10 PM
loscone South, Level 2 | San Francisco Ballroom 214

2. Acme Mart Demo
Cloud Integration Platform

uz)

Cloud ¿7
APIs

Messaging
a.
API App & Data Open API Ne
Ñ Management Integration ”
Mobile .. u...
“ ea
we a
=. Real-time "Order Management
a Events ya Systems

ug (CET
[om]
CES

3. ‘IBM is first-to-market with a holistic integration platform
exploiting container-based portable architecture for a range of
hybrid integration use cases, as well as providing essential
advice & support to help enterprises succeed with their
integration modernization initiatives.”

Saurabh Sharma
Principal Analyst
Ovum

4. « Supermarket chain based in
Spain
* Acquired a U.S. grocery
chain

« Wants to run promos based
on customer purchase habits

« Mainframe data on-
premises

|. Private cloud based on data
center in EU

Powored by

5. Challenges

Acquisition introduces
introducing another cloud
platform and data

They have tens of thousands
of data integration processes

They have constraints

» Cost for moving data out of
cloud platforms

« Processing data as close as
possible from their data
sources

« Regulatory constraints

6. Digital
Transformation

Hybrid, Multicloud
Adoption

More
than just

Big
Data...

89%

62%

Share of organizations
have adopted or have
plans to adopt a

“digital-first” strategy

CEOs/Sr.Execs have a
management initiative /
transformation program
to make business more
digital

94%

67%

Share of enterprise
clients using multiple
clouds

Share of enterprise
clients using

more than one public
cloud provider

7. Preferred Deployment is a mix of on-premise and cloud

Preferred Deployment Method

(N=209)
Cloud It’s critical that the tools be deployed
entirely via cloud
It’s our preference that the tools be One-third prefer a
deployed entirely via cloud mixed-cloud
We would prefer some mix of on-premise deployment

and cloud deployment
It’s our preference that the tools be
deployed entirely on our premises
It’s critical that the tools be deployed
entirely on our premises
No preference for how the tools would be
deployed

On-premise NEO

0% 20% 40% 60% 80% 100%

Which of the following best describes how your company would prefer that the Data Governance and Integration Platform be
deployed?

Powered by
deli 2018 / DOE W | Fahruser 2080 10 2048 Wed Canale eee

8. , sj
[BM’s Vision
= Smart and optimized data flows

+ Processing in close proximity to your data

+ Distribute processing to multiple clouds or on-prem
+ Optimize based on multiple criteria

Polyglot Execution Engines
+ Spark, IBM PX, IBM Streams, replication, Hadoop, etc..

= Smart Data Flow Designer Private Cloud
+ Intent driven, autonomous and self-learning (onPremises)
+ Smart data movement
+ Smart data replication
+ Smart ETL (transformations)
+ Batch, Near Real-Time and Real-Time

= Governance infused (data privacy and
protection)

= SaaS and PaaS

IBM Cloud

ray

9. Multi-cloud Data Integration Topology
Today

* Runs on-prem
+ Move data out of AWS for each
job execution

Frequently changing
table1 on prem

Staging table3 on AWS

Optimized flow

+ Same user design

* Flow execution has been moved to AWS

+» Remote table has been replaced with replicated tat

+ Generated replication flow
+ Anew staging table now available on AWS

10. Multi-cloud Data Integration Architecture

En

— Logical Flow

Intelligent Optimizer and Orchestrator

Private
AWS Azure IBM Cloud On-prem a

eee méme

Business Intent

data Integration and
movement

Jeployment

11. Multi-cloud Data Integration Architecture

En

— Logical Flow

Intelligent Optimizer and Orchestrator

Private
AWS Azure IBM Cloud On-prem a

eee méme

Business Intent

data Integration and
movement

Jeployment

12. Multi-cloud Data Integration Topology

Design Environment

Remote execution Remote execution

2 | executabe 1 je] :

IBM Cloud, AWS, Azure, On-prem IBM Cloud, AWS, Azure, On-premises

13. Key differentiators

Embedded
Governance

+ Enforce data
quality and
policies

Optimize for
your needs

* Cost, SLA, etc.

+ Distribute across multiple
clouds and runtimes

« Embedded machine
learning

Operational
Controls

* Single view of
operational
metadata- logs,
monitoring,
performance

14. The AI Ladder

4 prescriptive approach to accelerating your journey to AI

INFUSE — Automate and scale across your processes LS
TRUST - Achieve trust and transparency in outcomes
ANALYZE - Scale insights with Machine Learning everywhere
ORGANIZE - Create a trusted analytics foundation

COLLECT - Make data simple and accessible MODERNIZE
your data estate for an Al
and multicloud world

ei. ‘eo Data of every type, regardless
Ls of where it lives

15. IBM Data & AI Portfolio "

Everything you need for Enterprise AI, on any cloud

Build Run
The Ladder to Al Watson
Machine
SUR > Learning

Sn .

CD)

Open Source meets a ae:
multicloud, working as ONE 5 ee: tan

Infrastructure built for the most demanding workloads, infused with intelligence, with security
and data privacy protection
IBM Power Systems, IBM Z and LinuxONE, and IBM Storage

16. Adopting and Expanding AI

INGEST

A Single View of the Truth

Collect and normalize multiple data sources

+ Global requirements: IoT, Mobile, Sensors
Client data

Supply Chain Data

Transactional Systems

Client Behavior

Standard data analytics tools extract relevant data
+ ETL: Extract, Transformation and Load

+ Spark for real-time analytics

+ Scripted, repeatable, reliable and fast

IBM Chief Data Office
4 billion cient data
records

across

data source
refreshed week

Serving 29,000 employees
in 60 countries by YE18 *

17. Adopting and Expanding AI

Classify and Prepare Data Sets

ES]
A Training and Testing Data Sets
Scenes Tine

+ Accuracy improves with volume of data
Nes SREPARE + Track data sets to identify bias, create
£ related models
Make data available for other analytics
+ The Data Science Toolkit includes:
Hadoop, SPSS, SAS, R, etc.

Metadata Generation

+ Quality improvement and regulatory
compliance requires data tracking

+ Use Storage and AI to automate metadata

IBM Chief Data Office
Automated Metadata
Generation

90% reductio

in cycle time

Over 200k experimel
used to class

TBs of data
make it discoverat

18. Adopting and Expanding AI

Develop AI models

Increase Model Iterations
+ Fastest copy possible to GPUs

=
INGE?) a EOE + Most models can be distributed
Pé across multiple GPUs
a SOLE models for tracking and IBM Storage
E Ñ Fast, Scalab
e = Leading Edge Shared Data Service as , Cala

+ Leading GPU can cost $10k
+ GPUs: up to 150GB/s of throughput
+ Shared Container Service

Storage

2TB/s at COR
200GB/s in a single ra

19. Adopting and Expanding AI Storage B

Deploy AI models

— Low Latency APIs
NGEST Y PREPARE "Y TRAINING N INFERENCE + Fast storage for fastest
response
IBM Chef Data Office
=. Scalable Al service Quick & easy.
= SE +» Alis part of a portfolio of
applications <1 secon
* Containerized deployments response ti
to manage workloads 40,000 API calls/quart

Presented in familiaruser

TOM Watton M

interface.

20. Building AI projects in Silos is not Efficient

Every Business Unit A Common Data
Builds their Own Platform
o 0
® o
O
o
No single source of the truth Build on the common platform,
Every team is on their own and every team is more efficient

ik 2019 / 2388 / February 13, 2019 © 2019 IBM Corporation

21. Enterprise AI requires Shared Data

AI Data Workflow

El

Science Time

x
4| PREPARE
> <=>

(Gao cou object storage nen IO se

Common Enterprise Datá
Platform

* Single Source of Truth

+ Unified Data Access
Faster Cycle Times

+ Data Network Effect
accelerates Al Adoption

Containerized, Cloud Native
applications

22. ata Monetization:
uilding a Data and AI Backbone for the Enterprise

IBM is rapidly executing our AI transformation
Priority Solutions: Future:
Co-create

to envision, build
deploy, and scale

=
Optimization

AI Enterprise Data Architecture

Vorid-Class Infrastructure:

[

Private Cloud Hybrid Orchestration & Compute Public Cloud Cloud”A" | Cloud'N.

Built upon a shared, flexible storage architecture

ASE
cc

Inderpal Bhandari )
[BM Global

Chief Data Officer

zation: Building
kbone for

IBM Common Enterprise
Data Platform
+ IBM Spectrum Scale
Elastic Storage Server (ESS)
1.5PB of Flash
APB of disk
‘Supporting analytics, AI, anc
scripts with IBM Cloud
Private containers

Gp ensceciumsete

23. Enterprise AI requires Shared Data Ze Common Ehlerpn =

Data Platform
+ IBM Spectrum Scale
AI Data Workflow Elastic Storage Server (ESS)
1.5PB of Flash
4PB of disk

Supporting analytics, AI,
and scripts with IBM Cloud
Private containers

INGEST PREPARE TRAINING INFERENCE

20 million 7
contacts En

mn 3 Automated availability wi <1
With 10 | Metadata dat:

Generation ata

million

ecords | 90% queries second

| ; response time
reduction | '" 34,000 API
| in cycle time seconds calls/quarter
2 rather than i
| minutes/hours le

24. BM Storage
ads the way in
I infrastructure

IBM Systems Reference
Architecture for Al
IBM Watson ML
IBM Spectrum Computing
IBM Storage
BM Accelerated
Compute Platform
IBM Power Servers
IBM Spectrum Computing
IBM Spectrum Scale & ESS
IBM Storage for
Autonomous Driving
IBM Cloud Object Storage

IBM Spectrum Discover
IBM Spectrum Scale

Washington University St.
Louis School of Medicine
& the Vanderbilt
University Institute of
Imaging Science:

“From an IT perspective, we
were seeking technology that
could set new records for
speed, scalability, flexibility
and efficiency. It needed to
help us deal with data that is
growing fast, in volume, variety
& complexity across siloed
systems.

Xiaoyu Jiang, PhD, Research
Fellow at Vanderbilt University
Institute of Imaging Science

I8M Spectrum Scale

Academically, people talk
about fancy algorithms. But
in real life, how efficiently
the models run in
distributed environments is
critical. IE e X

IBM Elastic Storage Server

#1 & #2 Supercomputers
built for AI

Summit and Sierra are the
fastest computers in the wo
and purpose-built for AI
workloads.

Together, more than 44,
NVIDIA GPUs and 400
Petabytes of IBM Storage

* 2.5 TB/sec single stream IO!

*1 TB/sec 1MB sequential
read/write

* Single Node 16 GB/sec
sequential read/write

*50K creates/sec per shared
directory

*2.6 Million 32K file creates/s

25. Insurance Powered by Financial Services Cloud =

Unify the experience to unlock customer loyalty

Agent Console

Relationship Builder & Map E

[Actionable Policyholder & Producer
sicyholder 8 Household Profiles Insights Financial Services Cloud

Lightning App Builder Active Policies and Rollup

26. Best of both worlds

ITOps
Based on ITIL
Highly planned, structured

roadmap development and
release processes

Traditional IT, physical IT
assets, data centers

System of record: configuration
management database

Q

DevOps

Based on agile

Agile, integration, continuous
delivery

Cloud providers, virtual
assets, XaaS

System of reference: cloud
data management system

27. TBM Multicloud Manager

AWS Azure

us

EY)

IBM
IP ms m
15 638

nor

29. olution: Intelligent Finding Analytics

$— o. 19):

— Finding Fix Groups
Analytics + Best location fix
+ Fully automated review multiple vulnerabilities
+ Eliminates False Positives
& uninteresting results
+ Completes review in
seconds

Application
Analysis

30. \pplying Cognitive Computing to security vulnerability analysis

lachine learning with Intelligent Findings Analytics*

» Learned results

= 98.91% accurate in eliminating false
positives
* Minimize “unlikely attack scenarios”

= Provide fix group recommendations
that resolve multiple vulnerabilities

à. >_>
* Built on Watson Machine Learning UN a, =
+ Trained by IBM Security Experts Q

+ Fully automated review of scan findings

cad

+ Patents pending

31. ntelligent Findings Analytics: Real-World Results

sord fequired for manual reviews

existing development workflow À Z

Real-World Scan IFA

‚Applications Findings Vulnerabilities Groups

Application 1 12,480 1,057 35
BS ApplicatioN2 247,350 1,271 103

RN
- -
< | none 746,979 Maga 4
\ 5 N

32. Another Coverage issue Open Source Components

Over 2/3 of companies regularly use Open Source packages in their code

These packages are frequently the entry point for well-know attacks such
as Heartbleed, Ghost, and Shellshock

To prevent this, companies need to stay up-to-date on the Open Source
packages and versions they are using

To be most effective, Open Source scanning should be done along Li
static testing in the DevOps cycle

33. EE INIT Ns PIAL ew

1. Tightly Integrated into app sec testing

— With Static Analyzer (and OSA entitlement) automatic discovery of Open Source packages and identification of
vulnerable packages.

— Results and remediation recommendations automatically included in AsoC Portal and Static Analyzer reports

2. Uses industry’s leading sources of vulnerabilities

— Broadest, most up-to-date set of identified Open Source vulnerabilities from public and proprietary sources
+ Database of 3.5 million binary components and % billion source files
+ 300,000 identified vulnerable components

— Most competitors only use National Vulnerability Database (NVD)

— NVD has <10,000 Open Source vulnerabilities (10% of its total)

— OSA has an additional 11 sources of data beyond NVD

— Continuous monitoring for updates = continuously growing list

— Multiple sources of remediation advice, including from IBM X-Force
+ Includes links to patches, specific source files and newer versions that fix issues

3. Largest list of support languages

— Java, .NET, JavaScript, PHP, Node.JS, C/C++, Ruby, Python, ObjectiveC, SWIFT, Go, Scala, Clojure, Groovy,
Android, Perl, and Pascal

34. Achieving continuous security
nvolves an end-to-end solution

Manage Access Protect Data
Network
Threat
Protection
| Identity Data-in-use Data-at-rest
& Access Protection Encryption

—

Secure Platform

ink 2019 / © 2019 189 Corporation

35. AI-infused security insights

e Single pane of glass for security posture

À . Integrated vulnerability and certificates
Custom enterprise integrations

Open APIs and Partner Integrations

Network Insights*
with Security Advisor

CN EN

36. 3uilt-in security provide on-ramps

0 IBM Security add-ons

Manage Access
Managed & Identity Services
Consulting PAM as a Service
Services TAM Journey to Cloud
\dd-on IBM Security
[| Cloud Identity
Products Trusteer
Maas360

Oo Manage platform access w/ IAM

© Manage application access w/ App ID
|

Manage identity risk, governance...
unk 2019 © 2019 IB Corporation

© revProtect

© rer Protect crypto

Encryption & Key management

Security logs, events & flows

Powored by ;

37. comprehensive IBM Cloud Security Portfolio

38. Sybersecurity
Approach

Develop

KONE software
& devices

Collect & store data

KONE IoT
Cloud platform

9 18m Corporation

Equip

KONE
connected equipment

Provide safe & innovative
products & solutions

KONE solutions
& supporting services

Connect

Mobile network
connection

Detect & react

Maintain
& enhance

39. KONE & IBM Collaboration KONE

low we use IBM App ID for KONE Residential Flow

O

IBM Cloud Services
(App ID)

- Ease of use utilizing current
building blocks

- Integrated to enterprise
system

- Scalability - part of Active
Directory

40. ntegrate identity into multi-
loud apps

1 Built-in Identity Service, App ID

Simplified developer experience
User & App Authentication

_

User management & Profiles
— Open standards

K SIEMENS

Multi Factor Authentication
Multi-Cloud Policy Management*

in IBM Cloud Private

41. he untapped market potential
or tokenization is vast

19

Bitcoin or Equities Real Estate
0.06T 73T 217T

42. AUNCH OF Digital ASSET CUSTOUY service (UALS) cutie

dacs

\ /

E

Bespoke business logic for

custody transactions
\ Hyper Protect zZ
Virtual Servers 4
Corporate

o
a DR Audit

IBM Cloud }
Ze IBM Hyper = |
Protect a dE ZA m @
aye Banks Hedge Funds Exchanges
Services

te

oy,
—_) IBM Cloud Private