IBM presents:

Online live events 5-part series

All sessions start at 12pm Eastern / 9am Pacific

—
Part 1: April 21
Scale your business, empower creativity and drive innovation with
AI

Part 2: April 22
Expand AI talent: From data scientists to developers to analysts

Part 3: May 19
Monitor and govern models: From explainable AI to responsible AI

Part 4: June 1
Modernize your predictive and prescriptive analytics
Part 5: June 15
Extend your DevOps architecture and tools to be AI- and ModelOps-
ready

→ Register
ibm.biz/nextinAI
1

5-Part Data science and AI webinar series

What’s Next in AI

Part 4: Modernize your predictive and

prescriptive analytics
June 1, 2021

Gianmaria Leo
CPLEX Software Scientist & Developer, IBM Decision Optimization
IBM Data and AI
email: [email protected]/
Linkedin: https://www.linkedin.com/in/gianmaria-leo/

Julianna DeLua
SME, Platform lead, Data Science and Cloud Pak for Data
IBM Data and AI
email: [email protected]
Linkedin: https://www.linkedin.com/in/juliannadelua/

1 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

© 2020 IBM Corporation

 • As a reminder, this webinar will be available on-demand, and you’ll receive a link
within the next couple of days that will give you access to the archive.
• This webinar is being recorded and will be available on-Demand for one year.
• Today’s event will be interactive … so please disable any pop-up blockers you may
have installed -- they can interfere with certain aspects of our web conference.
• For an enhanced view of our presentation, click the “Enlarge Slides” button,
located in the gray bar above the slide window.
• To download a PDF version of the slide presentation, just click the Green Folder
icon located at the bottom of your console.
• You may ask a question at any time. To submit yours, please click on the Q&A icon
at the bottom of the screen and type it into the question box area that appears …
and then hit the submit button. We will try to get to as many questions

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

Please note

IBM’s statements regarding its plans, directions, and intent are subject to change
or withdrawal without notice and at IBM’s sole discretion.
Information regarding potential future products is intended to outline our general
product direction and it should not be relied on in making a purchasing decision.

The information mentioned regarding potential future products is not a commitment,

promise, or legal obligation to deliver any material, code or functionality. Information
about potential future products may not be incorporated into any contract.
The development, release, and timing of any future features or functionality described for
our products remains at our sole discretion.

Performance is based on measurements and projections using standard IBM benchmarks

in a controlled environment. The actual throughput or performance that any user will
experience will vary depending upon many factors, including considerations such as the
amount of multiprogramming in the user’s job stream, the I/O configuration, the storage
configuration, and the workload processed. Therefore, no assurance can be given that an
individual user will achieve results similar to those stated here.

4
IBM Cloud and Cognitive Software / © 2021 IBM Corporation

2 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Analytical approach

The 2021 CEO Study

3,000+
Chief Executive Officers

Analytical Approach
Customer-focused CEOs Product-focused CEOs Operations-focused CEOs Non-focused CEOs

48% 30% 17% 5%

Customer experience Product and service innovation Operations innovation Focus distributed across 2 or
innovation 3 aspects
Priorities: innovation in Priorities: improve efficiency,
Priorities: customer products and services, improve distribution channels,
relationships, customer scalability of service delivery, harden supply chain, reduce
experiences, new customer effectiveness of marketing and silo-to-silo handoffs, enhance
acquisition, end-to-end sales operational agility and flexibility
customer service delivery Technology impact/benefits:
Technology impact/benefits:
Technology impact/benefits: product and service innovations reduce costs, process
customer service, experience, automation
2020: building supply chain
insights, connect with new 2020: accelerating agility,
customers
capacity
reducing operating costs
2020: engaging customers
virtually

5 Source: IBM, April 2021, https://www.ibm.com/thought-leadership/institute-business-value/report/ceo

All respondents—Perform with purposeful agility

Enhanced operational agility and

flexibility is the clear top priority
on the next 2–3 years

Across all respondents,

operational agility and flexibility
is identified as a priority at least

56% 41% 40% 38% 37% 36%

37%
more than any other factor
Enhance Secure data Improve Establish Establish Reduce
operational and data greater flexible silo-to-silo
agility and systems transparenc accounta- partner handoffs
flexibility y end-to- bility networks
end

3 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Prescribe action by optimizing decisions on discovery and
predictions for modern apps
Discover and predict Optimize decision
Continuous
AI Model
Governance
Discovery
– Trends & patterns
Decision Optimization
– Correlations
– Scheduling
– Insight for hypotheses
– Demand and supply matching
– Allocation
Prediction – Logistics
– Segmentation and attrition – Pricing and category management
– Contact center optimization
– Scenario planning
– Predictive/preventive maintenance Multicloud
– Risk / governance / Fraud

7
To drive action and business results in a prescriptive way Containerized
Automated AI lifecycle management Apps

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

4 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 IBM Cloud and Cognitive Software / © 2021 IBM Corporation

IBM Data & AI

Modernize your Predictive &

Prescriptive Analytics

—
Gianmaria Leo, PhD
CPLEX Software Scientist, Lab Advocate
IBM Decision Optimization

10

5 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Agenda

1. A Use-case in Defense:
Building & Deploying Decision Optimization

2. Coupling Machine Learning & Decision Optimization:

A Project delivered in Financial Services

3. Tackling Uncertainty in Decision-Making:

A Solution delivered for Telco Services

4. Post COVID-19 Recovery:

Open-source contributions for Governments & More

5. Q&A

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

11

A Use-case in Defense:

Building & Deploying

Decision Optimization

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

12

6 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Optimizing Recruit Training of
a European Navy

• Typically, Navy plans missions to be executed

• Every mission has a release date to kick off operations

A navy mission requires:

• Equipment
(aircraft carriers, vessels, submarines)
• Specialized personnel with specific certifications
(engineers, nurses, navy specialists)

The recruit training offers:

• Designation and specialization certification programs
(CIS eng., medical assistant, submarine eng. technician)
• Courses to get one or more specific skills
(radio ops, drowning treatment, microminiature repair)

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

13

How does the training work?

The Navy organizes courses that have:

• Instructors (civilian and military)
• Facilities (equipment dedicated to the trainings only)
• Limited capacity (max number of allowed trainees)

Before jumping on any mission, recruits must get certified

for their designation (or specialization):
• Any certification is awarded when a given set of skills is
acquired by the recruit
• Any skill is acknowledged if any course covering the skill
has been completed by the recruit
• Some course cannot be taken until all its preparatory
courses have been completed (no chance to escape!)

Ready to join the mission?

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

14

7 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Complexities
Navy Training Use-case
Challenges
• Multiple certification • Computational performance
paths routing choices (due to the problem’s
• Multiple course nature)
sequencing choices
Goals, Complexity &
• Our client client built a
• Equipment and personnel solution based on a state-
availability of-the-art solver
Challenges •
•
Recruits’ time off
Yearly recruitment plan
• They were struggling with
unacceptable performance
against business goals
• Missions’ deployment • It took hours of
release and due dates computation to provide any
• Continuity of training feasible solution with poor
activities for recruits quality

Goals

• Optimal scheduling that preserves all the resource

constraints, operational rules and mission requirements
• The solution allows to target multiple KPIs, focusing on
both recruit training operations and Navy missions
• The user wants to interact with the optimization engine
• The user wants to evaluate quickly different scenarios by
changing dynamically inputs and obtained solutions

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

15

Results

• We investigated two
“Working with the team and using
approaches to model
the problem with CPLEX we were able to reduce the
time to find a feasible solution for
IBM CPLEX

Ø MIP engine

•
Ø CP engine

We were able to
this highly complex problem from
compute high quality
solutions in < 3 sec hours to seconds.”
• We deployed the
solution in hybrid
cloud environment,
with IBM Cloud Pak Vince Powell
Partner, Solutions Lead at Capita
for Data and
Watson Studio

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

16

8 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Design of the Prototype
An overview of the concept about the developed prototype and
its deployment as a cloud application

IBM Cloud Pak for Data

Data Handler IBM Decision Optimization Solution Handler

Model and solve with

Input data connector Output data connector
CPLEX CP

Get data via Model and solve with Expose solution via
REST API CPLEX MIP REST API

Client’s (Frontend) Application

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

17

Results – Base Scenario

Boat B Boat A

KPIs

Boat A: 86%
Boat B: 14%
Avg-Delay: 0.88
G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

18

9 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Results – Scenario 1: Change of Recruitment Plan
Boat B Boat A Description

All ETAB
available from
week 1

KPIs

Boat A: 100%
Boat B: 14%
Avg-Delay: 0.90
G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

19

Results – Scenario 2: Change of Facility Availability

Boat B Boat A Description

Facility 3 not
available during
weeks 11 – 13

KPIs

Boat A: 86%
Boat B: 14%
Avg-Delay: 1.94
G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

20

10 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Results – Scenario 3: Change of Platform Schedule
Boat A Boat B Description

Extended
deadlines for
Boat A and B
resp. to weeks
13 and 15

KPIs

Boat A: 100%
Boat B: 100%
Avg-Delay: 1.35
G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

21

Coupling Machine Learning &

Decision Optimization:

A Project delivered in
Financial Services

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

22

11 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 World ATM Market

$20.58 billions in 2019

… growing at

30.50 billions by 2027

at CAGR of 5.2%
from 2020 to 2027
*source:
https://www.alliedmarketresearch.com

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

23

Managing ATMs

• Outages are • ATM profiling to

common identify busiest
(out of cash, card machines on
reader fault, etc.) monthly/ daily/
hourly basis
• Modern ATMs have
sensors to enable • Resolve outages in
remote operations shortest time
while minimizing
• ATM uptime is operating cost
crucial
• Proactive
• Manage field maintenance
personnel to solve (before machine
incidents breaks) to reduce
outages

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

24

12 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Typical ATM Use-case Why? How?

Ensure ATMs are in Leverage data and

working conditions to exploit analytics
– Operations Manager needs to give the bank’s techniques and
prioritize jobs (incidents, predictive customers technology to reduce
maintenance, cash replenishments) the most pleasant costs, increase
and optimally assign field service banking experience profitability and take
personnel to perform daily jobs informed decisions

– Field Service Personnel needs to

know the sequence of jobs for ATMs
experiencing issues
– Area & Regional Manager needs to
monitor and control operational
(utilization) and financial KPIs of her
area/region

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

25

Overview of the Modeling

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

26

13 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Demo (Sample Prototype)

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

27

28

14 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 29

30

15 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 31

Tackling Uncertainty in
Decision-Making:

A Solution delivered
for Telco Services

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

32

16 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Achieving Cost Efficiency
Challenges Expected Benefits

• Projects have cross

for the Management of
• Increased awareness
dependencies, and their about risks and impact
execution is affected by of actions
Complex Telco Projects high uncertainty
• Increased accuracy of
• Activities are difficult to budget estimations
control due to complex
• Improved awareness
The Challenge of Tackling task relations

Frequent rescheduling
and control of third-
party dependencies

Uncertainty
•
impacts business results • Improved cost efficiency
and data quality

Goals

• Complex Telco projects focus on the mid/long-term

delivery of different workstreams that support the core
business operations of the company

• Workstreams have shared activities and resources

between different projects, complex dependencies and
different stakeholders with different KPIs and goals
• Decisions and mitigations often require a huge
collaboration and analysis effort before being applied,
within a remarkable fast-paced working environment
G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

33

Key Ideas & Results

“I am very satisfied about the
Ideas Results
project results and how my
• In this case, unbiased • Quantitative metrics to
team and IBM worked together.
datasets lead to
predictions with poor
measure risks and
assess the impact of
We were able to achieve
performance mitigations
valuable business
• Simulation techniques
help to build many
• Different views to
different stakeholders recommendations and
possible scenarios
then identify
on the top of
centralized modeling actionable insights by
unexpected risks
• PMs can quickly leveraging advanced Data
Robust Optimization identify criticalities
Science and AI techniques”
•
proposes decisions and risks
that remain feasible
and pursue project • Optimized
KPIs under the effect recommendations
of the uncertainty lead to costs
decreased by ~50%
Akram Dahy
Technical Team Leader at STCS
G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

34

17 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Solution Modeling
An overview of the pipeline deployed in IBM Cloud Pak for Data

Data User
defined
parameters
& requests

Machine Learning Simulation Decision Optimization

Preprocesse
d data
Probabilistic
predictions Deterministi Stochastic Robust
c Optimization Optimization
Optimization (heuristic) (exact)
Deterministi SOWs
SOWs Graph
c Scenario
Scenario
Scenario
s
predictions ss

Network Monte Carlo

Descriptive + Network Bi-level
Simulation Model based
Predictive statistics Optimization Programming
Reformulation heuristic

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

35

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

36

18 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Post COVID-19 Recovery:

Open-source contributions for

Governments & More

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

37

Executive Summary |
Emergent Alliance
Due to the current Covid-19 outbreak, we found it
critical to assess the Regional Status of an area to
help local authorities determine the level of risk to
health and the change of behaviour and sentiment
(pulse) in the population.

WHY? We aim to supply governments with insights which

could be used to support decision making of Covid-19
countermeasures, through the development of the
following outcomes:
• Level of risk to health associated with a region
• The pulse (sentiment and behaviour) associated
with a region
• Simulation of the Covid-19 social and economic
impact on the economy

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

38

19 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Emergent Alliance | Use Cases
“How to assess the Covid-19 risk on a regional level?”

Health

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

39

Emergent Alliance | Use Cases

“How people changed behaviour and sentiment during the Pandemic?”

Mobility

Sentiment

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

40

20 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Emergent Alliance | Use Cases
“Understand How and When we will economically recover from Covid-19”

Creation of analytical Proof of concept Emergent Economic

Economy model of the economy Engine app

Input-Output model with time dynamics Dashboards with use cases:

to simulate sectorial shocks and What if the transportation sector went
countermeasures into lockdown?

Cloud Pak for Data Cloud Pak for Data Cognos Analytics IBM Cloud Db2

Open source tools

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

41

COVID-19 Crisis
Response
A simple Watson Studio
demo showing how
Machine Learning &
Decision Optimization can
help Governments and
Health Authorities

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

42

21 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Optimal Allocation Policies for UNIQUE CHALLENGE EXPECTED BENEFIT

Healthcare Facilities & Services Implement data-driven

strategy to manage
Increase awareness
about capacity
and control the
Boost service level and
Allocate COVID-19 patients healthcare capacity
reduce costs
over the healthcare facilities Control the trade-off
Respond proactively to
between people safety
network to achieve safety and and cost efficiency crises, like pandemics
cost efficiency

PERSONAS USE CASE

Staff of the Ministry of Interior Predict COVID-19 patients' hospitalization in
Cabinet of the National Healthcare Authority Belgium provinces
Management of a Private Healthcare Group
Optimize the trade-off between hospital beds
saturation at province level, and patient transfers
DATA & TESTBED cost between provinces
Open data (gov) of Belgium
Analyze different scenarios and visualize solutions
G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

43

About the Asset

Predictive Prescriptive Viz

– Shows how to build fast
an end-to-end solution
with few development Optimize
Deliver
effort, while focusing Safety & Cost
insights and
more on the analytics Predict COVID-19 Efficiency
recomms
modeling (value) hospitalizations in controlling
through
Belgium provinces hospitalizatio
business
– Leverages open data ns and
KPIs
transfers
– Provide update results
on daily basis

– Combines Machine
Learning and Decision IBM Watson Studio IBM Watson Machine
Optimization models
Decision Optimization Learning
Experiments

IBM Cloud Object Storage

G. Leo / IBM Cloud / June 1, 2021 / © 2021 IBM Corporation

44

22 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Use Watson Studio to build trust throughout the AI lifecycle

IBM Research / May 19, 2021 / © 2021 IBM Corporation 45

45

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

46

23 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 How do you deal with disparate analytic tools
in various cloud or on-premise environments?

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

47

We have a data science team, but we don't

have optimization skills. what do we do?

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

48

24 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 how do you organizationally address to build
competency for predictive and prescriptive
analytics together?

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

49

Product Expert Advice Watson Studio Cloud Pak for Data Learning paths:
documentation trial Business Value Get started wth
Assessment Tool Watson Studio:

Learn more about Find an IBM Data and Try the interactive tool to Explore our demos,
Watson Studio benefits AI expert today for a Start at no cost receive a custom report product tours and hands
and capabilities free consultation to AI and the value and on-labs
savings of Cloud Pak for
ibm.com/analytics/ Data
Link expert-advice LINK ibm.biz/icpdbva LINK

For more information visit. Watson Studio site:

https://www.ibm.com/cloud/watson-studio

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

50

50

25 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 Easy onramp with Industry Client Attrition Dynamic
Segmentation
Accelerators: Ability to predict client attrition Advanced dynamic client
Out-of-box assets with at configurable points in the
future to protect revenue and
segmentation helps identify
unique cohorts of clients by
wallet share, while also building
– Sample data sets and schemas profound client loyalty
behaviors, account profile
– Jupyter notebooks to cleanse and prepare the data, run information, and
demographics
machine learning algorithms, and score the resulting models
– Sample dashboards to display the results
– API endpoints that you can call from other applications

Full list here:

Intelligent Streaming Analytics Life Event Prediction Offer Affinity

Maintenance
Perform continuous streaming Predict life and financial events Run campaigns effectively
Predict the lifespan and failure analytics to score customer impacting client’s financial lives by identifying client product
likelihood of a machine to event data using life event to help advisors proactively propensities or investment
determine the optimal prediction machine learning service a client’s needs theme affinities to drive new
or more desirable business
maintenance based on each models. This can be generalized
machines specific characteristic and applied to work with any opportunities
to avoiding unnecessary costs other AI/ML models.
associated with repair and
production loss.

IBM Cloud and Cognitive Software / © 2020 IBM Corporation

51

IBM is positioned
as a Leader in
Gartner Magic
Quadrant for Data
Science and
Machine Learning
Platforms*
*Gartner, Magic Quadrant for Data Science and Machine Learning Platforms, Peter Krensky, 1 March 2021 LINK
Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research
organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.

52

26 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 IBM receives highest scores in two use cases from Gartner
Critical Capabilities Report*

*Gartner, Critical Capabilities for Data Science and Machine Learning Platforms, Peter den Hamer, 13 March 2021
Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research
organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.

53

Data science use cases and patterns on Cloud Pak for Data
Address any use case and any data science pattern across any cloud

Use cases Customer experience Operational excellence Digital transformation Risk and governance Fraud
Industry Accelerators

Data science Multicloud data and AI

patterns AI lifecycle automation Decision Optimization / ModelOps / MLOps Explainable AI
and augmentation Intelligence deployment

Build and scale trusted AI. Unify predictive and prescriptive analytics. Empower users of all skills.

User persona ü Programmatic data scientists ü Visual /citizen data scientists ü Developers ü Analysts ü Subject matter experts ü Model governance lead

Activate your open source investments with Anaconda Repository

Data and AI Watson Studio (Build, Run and Manage)

platform
AutoAI , Open source tooling (Jupyter Labs, Jupyter Notebook & R Studio), Visual Data Science, Data Refinery, Decision Optimization, Model Execution in Hadoop

IBM Cloud Pak For Data

One Platform - Any Cloud

AWS Microsoft Google
Cloud IBM Cloud Private cloud
Azure
IBM Cloud and Cognitive Software / © 2021 IBM Corporation

54

27 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 IBM Cloud Pak for Data Explainable AI and model Reduced model monitoring
monitoring on IBM Cloud Pak for effort:
Data 35% to 50%
Reduced infrastructure
management effort Increased models in
65% to 85% production:
Between 3x and 8x

Increased accuracy
Data science, ML and AI of models:
benefits:
15% to 30%
$1.2 to $3.4 millions
Increased profits :
$4.1 to $15.6 millions
“We wanted AI and ML data capabilities that sit on-prem but still
have the modern cloud benefits. And by having a standardized “Our models are now more accurate, which means we can better
piece of infrastructure, with OpenShift being the orchestration forecast our required cash reserve requirements. A 1%
platform, Cloud Pak for Data checked all the boxes.” improvement in accuracy frees up millions of dollars for us to
lend or invest.”
- Managing director of artificial intelligence innovation, consulting
services industry Source: A Forrester New Technology: Projected Total Economic Impact™ Study Commissioned By IBM, February 2020
- Data scientist, financial services industry
55
Source: A Forrester New Technology: Projected Total Economic Impact™ Study Commissioned By IBM, August 2020
IBM Cloud and Cognitive Software / © 2021 IBM Corporation

55

Forrester Research named IBM a Leader in The Forrester Wave™: Multimodal

Predictive Analytics and Machine Learning (PAML), Q32020

IBM is packed with AI lifecycle services everywhere—public,

private, and on-prem.

What do you get when you combine a full stack of data analytics
capabilities —from data management to PAML to business intelligence —in
a microservices framework that can run seamlessly on-premises, in the
private cloud, and in multiple public clouds? IBM Cloud Pak for Data, an
offering that makes capabilities across the PAML lifecycle available when
and where your users need them. The crown jewel of Cloud Pak for Data is
Watson Studio, a PAML offering that combines easy-to-use, SPSS-inspired
workflow capabilities with open source ML libraries and notebook-based
interfaces. IBM continues to add innovations from IBM Research like
fairness monitoring, bias mitigation, AutoML, and federated learning.

IBM offers a compelling, scalable, increasingly integrated, and harmonized

platform with differentiated capabilities that spans the entire PAML
lifecycle and can be deployed anywhere across any cloud. Users will have
to navigate some technical previews and quickly evolving features, but that
will save them from otherwise having to stitch together a mismatch of
proprietary solutions and open source software.
Source: Forrester, The Forrester Wave™: Multimodal Predictive
Analytics And Machine Learning, Q3 2020
The Forrester Wave™ is copyrighted by Forrester Research, Inc. Forrester and Forrester Wave are trademarks of Forrester Research, Inc. The Forrester Wave is a graphical representation of Forrester's call on a market and is plotted using a detailed spreadsheet
with exposed scores, weightings, and comments. Forrester does not endorse any vendor, product, or service depicted in the Forrester Wave. Information is based on best available resources. Opinions reflect judgment at the time and are subject to change.

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

56

28 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 IBM Industry Leadership
IBM Watson Studio: IBM Watson IBM Cloud Pak
Multimodal Predictive Analytics Knowledge Catalog: for Data:
and Machine Learning Solutions Machine Learning Data Catalogs Enterprise Insight Platforms

Forrester Wave™
and Forrester New
Wave™ Leaders

AI Lifecycle Platforms

IBM #1 in AI Industry Reddot

Market Share Awards Design Awards

Full source information for all analyst reports on this page is contained in the speaker notes.
The Forrester Wave™ and Forrester New Wave™ are copyrighted by Forrester Research, Inc. Forrester, Forrester Wave™, and Forrester New Wave™ are trademarks of Forrester Research, Inc. The Forrester Wave™ and Forrester New Wave™ are graphical representations of Forrester's call on a market. Forrester Wave™ is
plotted using a detailed spreadsheet with exposed scores, weightings, and comments. Forrester does not endorse any vendor, product, or service depicted in the Forrester Wave™ or the Forrester New Wave™. Information is based on best available resources. Opinions reflect judgment at the time and are subject to change.

IBM Cloud and Cognitive Software / © 2021 IBM Corporation

57

IBM is positioned in the Leaders category in the 2020 IDC MarketScape for
Worldwide Advanced Machine Learning Software Platforms (October 2020)

The IDC MarketScape notes IBM’s Watson Anywhere approach

brings AI to wherever the data resides —across any cloud —to
help companies unearth hidden insights, automate processes,
and ultimately drive business performance.

The report highlights IBM’s capabilities to automate model

building with AutoAI, simplify deployment of trusted AI, and
calls out the delivery of prescriptive analytics capabilities for the
design and deployment of optimization models within IBM
Watson Studio as one of the key differentiators.

The IDC MarketScape report said, “As per IDC research,

customers are very satisfied with IBM's comprehensive support
for the end-to-end ML life cycle in combination with flexible
deployment and management options.”
Source: "IDC MarketScape: Worldwide Advanced Machine Learning Software
Platforms 2020 Vendor Assessment", By: Ritu Jyoti and David Schubmehl,
October 2020, IDC # US45358820

DC MarketScape vendor analysis model is designed to provide an overview of the competitive fitness of ICT suppliers in a given market. The research methodology utilizes a rigorous scoring methodology based on both qualitative and quantitative criteria that results in a single graphical illustration of each
vendor’s position within a given market. The Capabilities score measures vendor product, go-to-market and business execution in the short-term. The Strategy score measures alignment of vendor strategies with customer requirements in a 3-5-year timeframe. Vendor market share is represented by the size of
the circles. Vendor year-over-year growth rate relative to the given market is indicated by a plus, neutral or minus next to the vendor name.

58

29 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation

 IBM presents:
Online live events 5-part series

All sessions start at 12pm Eastern / 9am Pacific

—
Part 1: April 21
Scale your business, empower creativity and drive innovation with AI
Part 2: April 22
Expand AI talent: From data scientists to developers to analysts
Part 3: May 19
Monitor and govern models: From explainable AI to responsible AI
Part 4: June 1
Modernize your predictive and prescriptive analytics
Part 5: June 15
Extend your DevOps architecture and tools to be AI- and ModelOps-
ready

→ Register
ibm.biz/nextinAI
59

60

60

30 Group Name / DOC ID / Month XX, 2019 / © 2019 IBM Corporation