WHITE PAPER

ARTIFICIAL INTELLIGENCE AND

MACHINE LEARNING
Industry Insights and Applications

Abstract
Automation, Artificial Intelligence (AI) and Machine Learning (ML) are
pushing boundaries in the software and hardware industry to what
machines are capable of doing. From just being a figment of someone’s
imagination in sci-fi movies and novels, they have come a long way to
augmenting human potential (reducing risk of human errors) in doing
tasks faster, more accurate and with greater precision each time – driven
by technology, automation and innovation.
This is indeed creating new business opportunities and is acting as
a clear competitive differentiator that helps analyze hidden patterns
of data to derive possible insights. AI and ML can certainly enrich
our future thereby making the need for intelligent and sophisticated
systems more important than ever.
Automation comes to the core of this modernization where mundane
tasks are being automated using AI and ML. B2B and B2C companies
are adopting this technology and everyone is becoming a digital
technology enabler. This is predominantly a cultural shift that every
organization is going through and thereby elevating user experience.
External Document © 2019 Infosys Limited
CONTENTS

Abstract 01

Business and Technology Solutions using AI

and ML 04

Business/Technology Trends 05

Practical approach to AI and its

implementation 06

Benefits and approaches to AI 06

AI and its use cases 07

Conclusion 08

References 08

About the Authors 08

External Document © 2019 Infosys Limited

 Business and Technology Solutions using AI and ML
AI today is widely used and practically processing data and recommendations would mean nothing unless intelligence
applied to help business make informed to businesses to make decisions for very is loaded onto it – through programs and
decisions and improve customer specific problems– at a greater speed, scale software written by humans. Excessive
experience. AI is all about acting in a similar and velocity than what a human can do. automation can do more harm than
way, using algorithms as to what humans good, and humans are underrated in
The first challenge that exists in
would do, but in a much more controlled, that context; the true potential lies in not
implementing automation and AI is
efficient and faster manner. It also refers replacing humans but to augment and
identifying a use case. AI is all around us –
to information derived by applying a set amplify them.
ranging from unlocking our phones with
of algorithms with little or no human
face recognition to self-driving cars. A lot of us today have AI enabled systems
intervention.
in their homes such as Siri, Alexa and
Smart Home is an example of Artificial
Humans can put general reasoning, apply Google Assistant to name a few . AI is
Intelligence where home appliances are
situational actions and make informed becoming integral part of our daily lives
connected to each other and communicate
decisions much quicker. In simple terms, AI which will go on to influence in much
information that can help create alerts.
is the machine’s ability to think for itself. It wider terms, in everything that we do.
Even the best of AI machines needs to
aims to augment human capabilities and These are real-time interaction systems.
be controlled by humans, as a computer
not replace it. It is all about algorithms,

External Document © 2019 Infosys Limited

Business/Technology Trends
Most C-level executives either have already that is helping manage mind-boggling and pattern are best candidates for
adopted or have plans to adopt AI into volumes of data. The sensors and chips that automation. Lots of tools/techniques are
their organizations. The future of Artificial exist in physical things around us present available in the market that help achieve it.
Intelligence in IT is going to be a landmark a huge opportunity in tapping valuable Even a simple Unix script written that can
and will see an astronomical rise in the data, running analytics and using those we execute the task unattended is automation
growth of the following key emerging can make informed and better decisions in its simplest form. Oracle’s announcement
technologies: with. For example, Fitbit health, a fitness of Oracle 18c - autonomous / self-driving
mobile device leading to better customer database is another perfect example of
• Bots
experience. In the times we live in, the automation.
• Artificial Intelligence
power in our mobile device is much more
We must apply automation in everything
• Internet of Things (IoT) than the power in our homes, enabled by
we do. Mundane work must be done
AI performing human-like tasks with access
• Blockchain manually the first time only; the next time
to endless data.
• Automation around must be automated.
The use of AI, on the other hand, can be
• Cloud Computing Cloud Computing is gaining significant
completely biased on the data used to
traction in the IT industry with an
Bots are going to be the next buzzword train the AI enabled systems.
increasing number of companies planning
in the years to come. They are no longer a
Blockchain is another key technology to adopt cloud (SaaS, PaaS, IaaS, Public/
hype and many organizations have started
that will define the way transactions are Hybrid cloud) and has completely changed
to bring alive the idea of getting closer to
done, making it more simple, secure the way clients consume technology. More
human-less and automated interactions
and cost effective, while ensuring a and more organizations are investing
which are more rule-based AI enabled
seamlessly distributed (not copied) heavily in cloud computing to free
.Auditory chat-bots that are more speech
networked financial system. It is often bandwidth of their Information Technology
based are going to be following this. The
referred to digital gold as well and has resources in handling more complex/
key to a successful Chatbot is bringing in
built-in robustness. Blockchain can’t be business-oriented tasks, thereby leading
clarity to what is needed - plan to have a
controlled by a single entity; hence is to a better Return on Investment (ROI)
plan before building a Chatbot.
secure, has no single point of failure, is and Total cost of Ownership (TCO). Typical
Artificial intelligence(AI) is getting into transparent, public and decentralized. examples of this are Infrastructure and
virtually every technology that is helping AML (Anti Money Laundering) and KYC Datacenter outsourcing, Software on the
make systems take smarter decisions. (Know Your Customer) practices adopted cloud such as. Salesforce, Workday and
Building intelligent Apps becomes a key in blockchain are key contributors to its more are leading to lower operational
here. For example, building a smart ERP widespread acceptance. costs and better catalog-based services
assistant or a smart camera. model. Cloud has truly become one of
Automation is leading the way things are
the mainstream core technology on which
Internet of Things (IoT) is enabled by being done today – simple tasks which are
businesses run.
Artificial Intelligence / Machine Learning repeatable and with a defined behavior

External Document © 2019 Infosys Limited

 Practical approach to AI and its implementation
One of the pragmatic ways in which AI are repeatable and the efforts saved here mundane tasks can be done by robots and
is ruling the world is creating doubts in can be repurposed to other important humans can focus on business centric tasks
people’s s mind if it will eventually replace tasks. and other priorities.
humans. The fact however is that it will not;
AI is all about instructing a machine to We should embed intelligence into what
it will augment and amplify what humans
enable, repeatedly similar tasks, modify we do as routine work and where it
can do so that humans become more
its operation based on a particular event, requires humungous efforts to process
efficient. To give some perspective of this,
automating the processing of large large volumes of data that require
can a teacher be made redundant when
correlation to be built into multiple
the student does their job well based on amounts of data and helping humans
systems. Artificial Intelligence is there to
the instructions given by the teacher? make informed decisions by using insights
augment and not replace human potential.
presented by AI technologies.
For example; if a human does 100 things AI is all about strategy, technology is just
as part of the job, AI tends to replace those Automation is at the core of AI; it helps an enabler of that to drive insights and
20-25 things that can be automated and in saving precious human hours so that help with decision making.

Benefits and approaches to AI

AI has long term benefits in making things automated, processing large amounts of data at speed of the light, predicting outcomes using
trends and patterns processing large amount of data.

One of the best approaches to implement AI is based on the four pillars below:

Assess Select

Four
Pillars of AI

Foster Implement

First and foremost, we need to carefully premise), costs associated with the operationally run it.
assess and develop a comprehensive implementation, people cost and cost to
Finally, is managing and sustaining an AI
strategy around data, our needs, skill learn and sustain the systems.
plan with a comprehensive roadmap to
assessment, timelines, training change
Third, is creating an environment of identify priorities and modifying them on
management and TCO analysis.
learning, implementing the change, an ongoing basis. This is the key to growing
Second, we need to select the right monitoring the input/output, manage the capabilities and expanding horizons to
software, right platform (cloud / on- risks and training the teams needed to allow a new set of application and needs.

External Document © 2019 Infosys Limited

AI and its use cases
As part of the recent study conducted by Infosys, the diagram below depicts the percentage of respondents in each industry who have
already been experiencing disruptions due to AI technologies:

Industries already experiencing disruption due to AI technologies

65% Telecom & Communication

63%
60% Service Provider
54% 53% Banking & Insurance
51% 49% Oil & Gas
48%
Retail & Consumer Product Goods
Media & Entertainment
34% Healthcare & Life Sciences
Manufacturing & High-Tech
Travel, Hospitality & Transportation
Public Sector

AI can help us across industries and some analytics. AI can help tremendously to offer personalized policies and data
of its abilities are as listed below: here as consumer spending power and analytics and visualization and can
decisions have direct impact on retail help improve customer experience
• The Finance industry can greatly benefit
industry including logistics, transportation, dramatically. .
from fraud detection and identity theft
issues, do sales forecasting, use chat-bots, manufacturing, suppliers, marketing etc. Some of the practical use cases of
enable personalized offers, hedge fund Any sort of insights that can influence Artificial Intelligence, in a nutshell, are:
management, customer centricity by consumer spending will be a great boon
• D
 etect usage patterns and suggest
recommending unique/customer specific and can help retailers to take corrective
options
offerings and lead generation. AI can also actions using AI e.g. focus on geographies,
help banks detect margin improvement • M
 onitor infrastructure and recommend
seasons, under-performing stores, errors
based on health data
areas using Robotic Process Automation due to paperwork etc.
(RPA) that can enable bottom line growth. • 
Predict analytics for data security and
• The Education industry can use AI data theft
With the recent trends/incidents on
to revolutionize education delivery,
financial frauds, detecting such incidents • Manage entire systems independently
personalizing and customizing based
becomes all the more important for
on student needs including global • Manage knowledge base
compliance and best governance.
classrooms, checking and grading
• 
Do sales forecasting and lead
• The Healthcare industry can be students, reviewing homework, generation
benefitted by analyzing tests and recommending higher education
providing personal treatment, advice institutes based on area of interest and • Perform customer service using bots
to patients using assistants/chatbots, past educational exposure/experience • Assist HR analytics
monitor and recommend based on and recommending learning paths.
• 
Perform data management, processing
individual health data. AI can also be
• The Insurance industry can use AI to and monitoring
used to medically train people and
detect frauds, process claims, suggest • 
Perform social media monitoring and
augment capabilities of human reach.
better policies and competitive policy marketing analytics
• In the Retail industry, chat-bots can help pricing, risk management, enable
• 
Detect system anomalies - Anti-Money
consumers interact with your brand more personalized offers based on health
Laundering
efficiently. . For example, detect usage status. This can be augmented by
patterns and suggest options, perform processing data from multiple systems • 
Predict technology failures - disk, file-
social media monitoring and marketing including Internet of Things (IoT) devices system, reactive analysis

External Document © 2019 Infosys Limited

Conclusion
The key how quickly and smartly we can adapt, adopt and embrace these changes, which will mark our organization’s and our clients’ success
in the years to come.

We look forward to ‘riding the wave’ with our clients from across the globe.

It is time to embrace AI in a controlled manner and ensure we excel in enabling our clients adapt to this technology through innovations we
bring to the fore.

Exciting times ahead indeed and the day will come when even technology books will be written by AI based on topical and current context
and events.

About the authors

Rohit Mohindru
Group Project Manager, Infosys

Rohit is a Group Project Manager at Infosys, leading engagements on Oracle technologies and providing leadership, guidance,
mentorship to teams.

Prabhat Kohli
Senior Technology Architect, Infosys

Prabhat is a Database Administrator and Senior Technology Architect at Infosys, helping clients align their Digital journey –
he is pioneering initiatives on automation and optimizations. He has extensive experience working on varied technologies
ranging from infrastructure, software, consulting, performance tuning and automation amongst others.

References
• https://www.infosys.com/age-of-ai/Documents/age-of-ai-infosys-research-report.pdf

For more information, contact askus@infosys.com

© 2019 Infosys Limited, Bengaluru, India. All Rights Reserved. Infosys believes the information in this document is accurate as of its publication date; such information is subject to change without notice. Infosys
acknowledges the proprietary rights of other companies to the trademarks, product names and such other intellectual property rights mentioned in this document. Except as expressly permitted, neither this
documentation nor any part of it may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, printing, photocopying, recording or otherwise, without the
prior permission of Infosys Limited and/ or any named intellectual property rights holders under this document.

Infosys.com | NYSE: INFY Stay Connected