For the digital enterprise, architecture of all varieties must evolve strategically in step with technological capabilities and business imperatives. Such a multidimensional approach includes automation, AI, analytics, big data management and digitization as a holistic phenomenon.
Notes on an ITO Appliance Approach to Productising and Industrialising IT Out...
Report
Share
1 of 11
Download to read offline
More Related Content
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning
1. Digital Enterprise
Architecture:
Four Elements
Critical to
Solution
Envisioning
Today’s digital organization
demands an enterprise
architecture that is guided by
its intended business outcome
and which can inform strategy
embracing a multidimensional
approach covering digitization,
data management, analytics,
AI and automation. Here’s how
to get started.
Cognizant 20-20 Insights | June 2018
COGNIZANT 20-20 INSIGHTS
2. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 2
Executive Summary
The digital revolution continues to transform our
world. Mobility is rapidly ascending as sensor and
actuator capabilities bring intelligence to devices
of all varieties via the aptly named Internet of
Things (IoT). According to IDC, our digital world
is now “doubling in size every two years, and by
2020 the digital universe – the data we create
and copy annually – will reach 44 zettabytes, or
44 trillion gigabytes.”1
As data proliferates expo-
nentially, a new class of applications is emerging,
one endowed with the intelligence to redefine
the business, operating and technology models
in place since the onset of the 21st
century.
Companies such as Google, Amazon, Facebook,
LinkedIn and Uber are leading the way in mon-
etizing big data and disrupting markets through
data-driven strategies. In this regard, McKinsey2
has emphasized the importance of an integrated
approach to data sourcing, model building and
organizational transformation.
The pivot to digital depends on a suite of tech-
nology developments in the areas of sensors,
actuators (for triggering actions), networking/
integration and computing (data management,
processing, analytics, etc.). This white paper
outlines the key architectural elements for under-
taking an integrated design approach and thereby
accelerating the journey to full-scale digital. The
foundational aspects of our approach are guided
by TOGAF,3
the industry-leading enterprise
architecture (EA) framework. The proposed archi-
tecture elements focus on what we call the four
M’s – materials, machines, models and mesh.4
THE OVERARCHING
PERSPECTIVE OF THE DIGITAL
ENTERPRISE ARCHITECTURE
The business world is increasingly interconnected
via constituent nodes of networked computers.
These nodes pervade business, IT systems and
applications of varying size, individual devices
and even sensors.
The digital architecture of a connected environ-
ment forms the foundation of an information
ecosystem that provides business services
to customers, business partners and employ-
ees. Such business services are composed of
finer-grained constituent services and data
from other nodes – which may be either within
or outside the business unit’s boundaries. New
services are formed by slicing, consolidating
The digital architecture of a connected
environment forms the foundation of
an information ecosystem that provides
business services to customers, business
partners and employees.
3. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 3
and repurposing information contained within
the extended enterprise and then, by applying
analytical and processing intelligence, generat-
ing new services of interest.
As a result, the functional logic, information
used, system and technology involved often
cut across applications, application architec-
tures and even enterprise boundaries, which are
transparent to the consuming user. Therefore,
a digital architecture blueprint is vital for suc-
cessful implementation of a digital enterprise.
However, service-level architectural parameters
such as uptimes and response times remain tied
to services in a real-time service integration
scenario, and need to be factored in when the
architecture is defined.
Digital Vehicular Traffic System: Illustrates Connected Ecosystem
Multiplying the Digital Enterprise Architecture Complexity
Traffic
Control
Vehicle
Insurance
Vehicle
Law
Enforcement
Accident
Management
Traffic
Signals
Actuator
Detector
Infotainment
Location
Warranty
Outage
Diagnostic
Parts
Supplier
Passenger
Subscriptions
Payments
Medical
Assistance
Servicing
Health
Insurance
Tolls
Parking
Weather
Data
Figure 1
Digital Architecture Blueprint: Four Focus Areas
ENTERPRISE DIGITAL ARCHITECTURE
Business Architecture
Focus on business-
outcome-driven models
and process views.
Information
Architecture
Focus on identifying and
creating a mesh of data
and managing the same.
Application Architecture
Focus on building
intelligent machines,
embodying the logic to
work upon the data to do,
learn, think.
Technology Architecture
Focus on the material
aspect of the intelligent
machines covering the
hardware and technology
choices of implementation
(e.g., AI software).
Figure 2
4. Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 4
BUILDING A BUSINESS
ARCHITECTURE
In our digital age, organizations need new models
and services to generate greater business value.
For this, they need to use the wealth of digital
data that surrounds their organizations – from
their people, processes, devices and consumer
input. We call these Code Halos,5
unique virtual
identities produced by every digital click, swipe,
like, buy, comment and search. Enterprises need
to be adept in digitizing, analyzing data and auto-
mating smart actions to attain an edge over rivals
and reap enhanced operational efficiencies in the
hyper-competitive global economy. This demands
new business solutions and approaches.
In our book What to Do When Machines Do
Everything, the authors from our Center for the
Future of Work outline a data-driven approach
(see Figure 3) for building intelligent machines
needed to thrive in the digital economy.
Intelligent process automation (IPA) applies the
machine intelligence (MI) baked into the algo-
rithms that power today’s software for creating
sophisticated business processes. IPA applica-
tion can be found in clinical data management
for life sciences, claims adjudication for insurers,
loan applications in banking, logistics optimiza-
tion, and traditional technology processes such
as infrastructure and information management
services. Sites recommending items based on
previous purchases are using MI to analyze users’
buying patterns and promote other items of likely
interest to the customer.
Enterprises need to be adept in digitizing,
analyzing data and automating smart actions
to attain an edge over rivals and reap enhanced
operational efficiencies in the hyper-competitive
global economy. This demands new business
solutions and approaches.
Getting AHEAD with
Business Architecture
utomation
Focus on unleashing the power
of the digital workforce comprised
of software robots (bots) and
virtual personal assistants (VPAs).
alos
Focus on the true power of
personalized services by cultivat-
ing information that surrounds
people, the organization and
devices.
nhancement
Focus on enhancing end-users’
perceptions of system aspects
such as utility, ease of use and
efficiency.
bundance
Focus on vast new markets by
dropping the price point of
existing offers and finding new
energy sources with the help of
machines.
iscovery
Focus on conceiving new prod-
ucts, new services and new
industries by leveraging machine
intelligence.
Figure 3
5. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 5
Enterprise digital architecture helps IT organiza-
tions to realize diverse business use cases such
as the following:
• Enhancing public safety and optimizing utili-
ties through sensor data analysis.
• Assessing patients’ health in real time by ana-
lyzing patient data from wearable devices and
sensors.
• Identifying fraud and money laundering via
banking cyber-surveillance.
• Route planning based on crowd-sourced GPS
data.
• Discovering new energy sources in the oil and
gas industry, and streamlining oil distribution.
With the advancement of smart meters, delivery
optimization and algorithmic trading, today’s
new machines are progressively leading to the
next generation of inventions. So the underlying
architecture needs to provide an extensible foun-
dation to support such evolution.
DESIGNING AN APPLICATION
ARCHITECTURE
In fact, digital advancement is progressing
much faster than what many industry watchers
expected. Machines can now read, see, listen,
write and actuate. Using these core capabilities,
machines can now perform complex tasks and
learn new things. Sooner or later, machines will
have the capacity to imitate the other human
senses: touch, smell and taste. And with time
they will be able to work smarter.
The challenge, therefore, for an enterprise appli-
cation architect is building intelligent systems that
can fulfill business goals which deliver results that
exceed competitors’ efforts and create additional
value for the customer. The architecture must
leverage the core capabilities available in some
distributed models and enable tighter human-ma-
chine collaboration than ever before.
This calls for a focused, methodical approach
for identifying how machines can advance auto-
mation, productivity and discovery by creating
and analyzing Code Halos generated through
the abundance of data that permeates business.
A retail organization, for example, can lever-
age four key areas under Code Halos where
In fact, digital advancement is
progressing much faster than what
many industry watchers expected.
Machines can now read, see, listen,
write and actuate. Using these core
capabilities, machines can now perform
complex tasks and learn new things.
6. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 6
machines can be a differentiating factor: iden-
tification of market segmentation, sentiment
analysis, campaign management and recommen-
dation engines. Similarly, there are four areas
of enhancement: predictive maintenance, pre-
dictive planning, cognitive monitoring and fraud
detection. (How to do it is a topic of separate
deliberation.)
Along with identifying these areas, common ana-
lytical practices (such as forecasting, optimizing
and planning), are required to build machines
in each respective area. Figure 4 illustrates
one approach. This design customizes machine
learning algorithms and models for common API
realization. As organizations shift to digital, fore-
casting, optimizing and planning activities shift
to real-time processes and become part of a con-
tinuous cycle rather than the periodic events of
yesteryear.
Architects need to identify the sources from
which they can acquire the relevant data
to model the enterprise as realistically as
possible. Along with identifying the sources,
they must identify or build interfaces that
can capture an information object with its
attributes.
Designing a Modern Digital Applications Architecture
Automation Halos Enhancement Abundance Discovery
Do
Learn
Think
Application Capability
Focus on building solutions and frameworks that can perform, learn and discover by themselves
and fulfill business objectives identified under the bucket of automation, Code Halos, enhance-
ment, abundance and discovery.
Analytical Capability
Focus on prediction, reporting, planning, optimization and collaboration algorithms and models to
build an end-to-end analytical layer.
Sensory Actuation Capability
Focus on creating core capabilities that imitate human senses such as reading, watching and listen-
ing to digitize the enterprise and generate data required to inform machine algorithms. Also focus
on triggering actions like toggling lights (e.g., of signals), physical movements (e.g., a motor), etc.
Figure 4
7. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 7
INTUITING INFORMATION
ARCHITECTURE
Information is digital’s key substrate, the base
upon which application capabilities and analyti-
cal capabilities thrive.
The challenge to enhanced digitization of the
physical business world starts with generating a
model that properly simulates the business. Many
organizations have begun to do this with their IoT
deployments and related digital initiatives. Archi-
tects need to identify the sources from which
they can acquire the relevant data to model the
enterprise as realistically as possible. Along with
identifying the sources, they must identify or
build interfaces that can capture an information
object with its attributes. Thus far, organizations
have been focusing primarily on objects and their
attributes, as success for a digital solution lies in
the ability to capture the holistic behavior of an
object. Moreover, with the proliferation of data,
the challenge becomes one of petadata manage-
ment – how to acquire, store and organize vast
quantities of quality data. When there is availabil-
ity of data aplenty, categorizing the information
is key to better management.
Finally, siloed data cannot reveal the full story.
The dots need to be connected to complete the
narrative. Hence, an integrated view of informa-
tion must be built by correlating the objects and
building a comprehensive enterprise informa-
tion model and associated interfaces. (A detailed
approach to achieve this is a discussion topic for
the future.)
Eyeing the Right Information Architecture
Integrated View
Focus on connecting information to build a real-world model that makes the information
livelier so that it all tells a better story.
Data Type
Focus on classifying data, whether it is structured, unstructured, audio, image, video,
document or sensor data coming from RDBMS, No-SQL, HDFS, CMS, etc., so that it can be
better managed when acquiring, storing, organizing and analyzing high volumes of data.
Interface
Focus on data from a diverse set of interfaces (e.g., sensors, telemetry, wearables, audio,
video, chat, e-mail) that not only capture attributes of a specific object but also its associ-
ated functionalities.
Source
Focus on sourcing data from enterprise systems, social media, mobile platforms, sensor
networks, print media, etc. so that anything and everything associated with the organiza-
tion can be digitized.
Figure 5
8. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 8
ENVISIONING THE TECHNOLOGY
ARCHITECTURE
Companies need to keep digging for digital
fuel (i.e., mining data of all types and formats)
to continually grow their business intelligence.
Special technologies and materials are required
to discover, acquire, organize and analyze big
data. Traditional technologies, such as CPU-
based computing and relational databases, fall
short in managing the volume, velocity and vari-
ety of data.
At the same time, technologies that power MI
are proliferating. A disciplined approach to find-
ing and applying the most appropriate tools is
critical. In this regard, an effective technology
architecture can ensure that this critical ele-
ment is in place at the right time in tandem with
addressing the key human and organizational
issues involved in a cultural change. Figure 6
summarizes key technical areas and their char-
acteristics.
IoT devices and platforms are crucial for data
acquisition and real-time ingestion of different
types of data. Once acquired, various data stor-
age technologies (i.e., RDBMS, No-SQL, CMS,
HDFS), along with data warehouse tools, help
to organize the data through transformation,
normalization, encoding, generating training
sets, etc. The processed data is fed to analysis
tools with built-in algorithms such as clustering,
learning, etc., and models such as predictive,
optimization, planning, etc. Finally, an enter-
Focusing on Technology Architecture
Business Solutions
Focus on leveraging service providers or commercial-off-the-shelf solutions. As machine
learning gains momentum, more frameworks will be packaged with solutions.
Technology for Analyzing Data
Focus on core technology like data science, machine learning and natural language
processing for off-the-shelf algorithms, models and different types of data analysis
capabilities.
Technology for Organizing Data
Focus on data capture and enrichment tools to organize data so that it can be ingested
into the analysis framework.
Technology for Acquiring Data
Focus on cognitive IoT sensors or devices for acquiring and disseminating structured and
unstructured information.
Hardware
Focus on compute and storage power required to handle big data; MI algorithms work
a lot more effectively on GPUs. Examples of such hardware are Google TPU, Cirrascale,
NVIDIA, DGX-1, Titan X, etc.
Figure 6
9. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 9
prise should appropriately experiment with
emerging industry-specific business solutions6
or MI platforms.
LOOKING FORWARD
The digital world has brought tremendous
opportunities and new challenges. The key is
to prioritize and convert the opportunities in
a systematic and holistic enterprise-focused
approach to create a digital solution. The solu-
tion needs to be functionally effective, practical,
supportive for a plug-and-play architecture and,
importantly, scalable in part and as a whole. The
solution design would typically focus on mod-
ernization and hardening of legacy and core IT
supporting multimodal architecture models. A
digital-ready, scalable architecture would prior-
itize specific aspects of architecture like cloud
adoption and security improvements focusing
on the four M’s of digital engineering discussed
before. It’s essential to avoid looking at it as a
narrow data, automation, integration, IoT or
analytics problem.
The solution designer needs to examine individ-
ual aspects of data procurement, gain business
insight and then convert this knowledge into
services that add business value. Each of these
aspects requires deep dives and elaboration
into architectural building blocks – starting with
technical capability. Dimensions such as security,
architectural governance and data stewardship
will then need to be addressed. Special attention
is required to manage the sensitivities related to
automation’s impact on people. Each issue could
be a topic for a separate in-depth deliberation.
FOOTNOTES
1 “Data Growth, Business Opportunities, and the IT Imperatives,” IDC, April 2014, https://www.emc.com/leadership/digital-uni-
verse/2014iview/executive-summary.htm.
2 “Three keys to building a data-driven strategy,” McKinsey, March 2013, http://www.mckinsey.com/business-functions/digi-
tal-mckinsey/our-insights/three-keys-to-building-a-data-driven-strategy.
3 TOGAF, http://www.opengroup.org/subjectareas/enterprise/togaf.
4 Malcolm Frank, Paul Roehrig and Ben Pring, What to Do When Machines Do Everything, John Wiley Sons,
February 2017, https://www.wiley.com/en-us/What+To+Do+When+Machines+Do+Everything%3A+How+to+Get+A-
head+in+a+World+of+AI%2C+Algorithms%2C+Bots%2C+and+Big+Data-p-9781119278665.
5 Malcolm Frank, Paul Roehrig and Ben Pring, Code Halos: How the Digital Lives of People, Things, and Organizations Are
Changing the Rules of Business, John Wiley Sons, April 2014, https://www.wiley.com/en-us/Code+Halos%3A+How+the+Dig-
ital+Lives+of+People%2C+Things%2C+and+Organizations+are+Changing+the+Rules+of+Business-p-9781118862070.
6 Shivon Zilis, James Cham, “The current state of machine intelligence 3.0,” November 2016, https://www.oreilly.com/ideas/
the-current-state-of-machine-intelligence-3-0.
10. Cognizant 20-20 Insights
Digital Enterprise Architecture: Four Elements Critical to Solution Envisioning | 10
Abhik Sengupta
IT and Digital Enterprise
Architecture Consultant,
Cognizant Consulting
Kamales Mandal
Enterprise Architect,
Cognizant Consulting
Tapodhan Sen
Architect, Cognizant’s Global
Technology Office
Abhik Sengupta is an IT and Digital Enterprise Architecture Con-
sultant within Cognizant Consulting. With nearly three decades of
industry experience, Abhik leads a team of enterprise architects
and legacy transformation specialists guiding and supporting major
enterprises across the world in their IT and enterprise architec-
ture strategy and organization transformation initiatives. He has
authored papers published in IEEE and other publications. Abhik can
be reached at Abhik.Sengupta@cognizant.com.
Kamales Mandal is an Enterprise Architect within Cognizant Con-
sulting. He has over 17 years of experience in diverse walks of
enterprise application development, enterprise integration and IT
consulting, and has worked with major enterprises across North
America, Australia, UK and Latin America in the retail, govern-
ment, healthcare and financial domains. Kamales has authored
multiple papers and is the key architect of an emerging semantic
technology-based, model-driven consulting platform: ACE (ana-
lyze. consult. execute). He can be reached at Kamales.Mandal@
cognizant.com.
Tapodhan Sen is an Architect in Cognizant’s Global Technology
Office. He has over 20 years of experience architecting complex
adaptive systems, focusing on interoperability and automation.
Tapodhan is an experienced enterprise architect with wide-rang-
ing experience at established blue chip companies and broad
exposure to various industry sectors including government,
finance, telecommunications, publishing and information media.
He can be reached at Tapodhan.Sen@cognizant.com.
ABOUT THE AUTHORS