CONVERSATIONAL AI 101

Conversational AI (CAI) Trends

Over the last few decades, much has been said about how AI will transform
virtually every industry by automating physical, transactional, and cognitive
tasks. And since the 1990s, with the advent of interactive voice response
(IVR), technology has been used to allow humans to interact with computers
through the medium they are most comfortable with: conversation.

These earlier attempts at allowing humans to "converse"

with computers left a lot to be desired. Conversations
with older IVR systems had to follow the strict structured
rules that often led users to a dead end and frustration.
But this technology has evolved and barely resembles
systems of the past.

Modern conversational AI systems can adapt quickly to

the nuances of a conversation. The current generation of
AI-enabled virtual assistants can recognize a wider
variety of user inputs, including those they were not
specifically trained on. Each new conversation becomes
a "learning experience" for the assistant as it integrates
the interactions and outcomes of a conversation into its
models. These interactions become even more powerful
when they are coupled with integrated applications
allowing users to trigger backend processes with only a
few words.

The popularity of Alexa, Siri, and Google Assistant has

proven the merit of modern conversational AI. According
to eMarketer, over 40% of internet users use a voice
assistant at least monthly, and NPR reported that 52% of
smart speaker users use their device on a daily basis.

These successful implementations of conversational AI in

all-purpose digital assistants have made many industries
take notice. The bridge between these conversational
digital assistants and virtual assistants ready for the
enterprise has not been the easiest to cross. We're not
talking about simple chatbots. In specialized business
environments, these digital assistants must not only have
the ability to recognize speech but also intent. Even so,
recent news is promising. Conversational AI is already
changing a wide variety of businesses.

CONVERSATIONAL AI 101 2 2
Here are some numbers worth noting:

“ More than 50% of

enterprises will spend
more per year on bots
“ 90% of all
customer interactions
with banks will be
“ The top technology
investment for insurers
in 2022 will be in AI. 74%

“
and chatbot creation than automated by 2022.
of insurers will be
on traditional mobile app
investing.
development.

“ “

“ 90% of the hospitals

in the US will use AI and
chatbots to save lives and
“ By 2023, over 70%
of all chatbots accessed
will be retail-based.
“ 63.5% of telecom
organizations are making
new technology
improve the quality of
care by 2025. “ investments for AI
systems.

“ “

“ The travel and

hospitality AI market is
estimated to surpass
the $1.2 billion mark
by 2026
“ Global AI use in the
media and entertainment
market is expected to grow
from $1.8 billion in 2018 to
$8.4 billion in 2023

“ “

Conversation AI is not only becoming mainstream in every

major vertical but a requirement to remain competitive. We
are past the early adopter phase.

CONVERSATIONAL AI 101 3 2
What is Conversational AI?
Conversational AI is the design and orchestration of a digital assistant that
resolves simple queries and questions without requiring human involvement.

A basic definition of conversational AI is the

design and orchestration of a digital assistant
that resolves simple queries and questions
without requiring human involvement.
Integrated intelligent virtual assistants (IVAs)
can extract data, inform, and resolve user
questions in the majority of cases passing it
on to a human only if the query is complex.
That, in turn, frees up employees and human
agents to handle more challenging tasks.

How Conversational AI works

The first time a neural network was

applied to a real-world problem was in
1959. Needless to say, machine
learning technology has advanced
considerably since then. Today's
generation of artificial intelligence has
evolved through decades of research
and experimentation.

The key to the CAI process is understanding the natural language of the user. This is called natural
language processing or NLP. There are many ways to implement NLP. More advanced systems like
Conversational AI Platforms (CAIP) process data through multiple natural language understanding
(NLU) and conversation engines to ensure the best match between intent and action.

CONVERSATIONAL AI 101 4 2
 The first step in the process is recognizing the words, phrases,
and sentences of users. What this step entails depends on
whether a user is talking on the phone or communicating via
text. For the NLP algorithms to understand voice data, it first
must be converted into text. Voice recognition algorithms
account for multiple languages, accents, and quirks that occur
in human speech.
Language
Recognition The user's utterance is then analyzed to identify each word by
its meaning, position, conjunction, plurality, and other factors.
Sentences as a whole are analyzed to determine the
relationship between words. And then, the final meaning is
determined by combining these analyses.

Determining what a user has typed or spoken is only the first

step. The next is to understand what the user is trying to
achieve. Many intents will consist of a simple verb and noun
combinations like "Find product," "Book a flight," or "Transfer
money." The user may also use modifiers like "then" or "first" in
the case of a phrase like "Book a flight then find a rental car."
Intent Multi-intent detection can identify the multiple action items a
Recognition user is describing and then execute their tasks in a logical order.

An entity is anything that defines, shapes, or modifies the user's

intent and is required to carry it out, such as dates, times, and
locations. When a user asks the conversational agent to book a
flight, it will most likely include a city, date, and time that must be
accounted for.

Entity Entity extraction identifies these elements in the user's words.

Extraction They can vary from simple numbers and dates to complex items,
such as addresses and airport names, to categories defined by
the user.

CONVERSATIONAL AI 101 5 2
 Why Conversational AI?
CAI enhances customer relations, improves
cost efficiency, provides better customer
feedback, brings you better insights, and
features uninterrupted service.

Sometimes the buzzwords and complex technology

distract from what conversational intelligence can do
for business when what you want to know is what
conversational AI can do for you.

Enhances Customer Improves Cost Better Feedback

Relations Management Efficiency

Conversational AI can minimize
the wait time of callers, can
help them with many issues
without human intervention,
and frees up the time of
agents.
CAI can serve customers
continuously, hone
problem-solving skills over
time, and cost nothing more
than some minimal
maintenance costs.
Conversational AI can elicit
feedback from users at various
points in their journey. And this
feedback can be used to
design more tailored offers
and promotions and inspire
more loyalty in customers.

Better Insights Uninterrupted Service Efficiency

CAI can access a wide variety of data that can be Conversational AI assistants don't have to take a
used to explore customer data. While human break or leave at the end of a shift. They work
agents may get to know specific customers over 24/7 for your business.
time and use their knowledge to better serve those
customers, a conversational AI system starts
gathering insights from day one.

CONVERSATIONAL AI 101 6 2
Virtual Assistants as a Manifestation of
Conversational AI
Your customers and employees can interact with digital systems the same
way they already communicate with other people or a brand.
The goal of a conversational AI system is to interact with employees and customers in a medium they are
comfortable with. These interactions happen through virtual assistants that can be customized in various
ways depending on what type of system you are using.
Conversational AI systems support an omnichannel approach to digital assistants. What this means is your
customers and employees can interact with digital systems the same way they already communicate with
other people or a brand or, for that matter in a language they are comfortable in. Instead of building a
virtual assistant for each one of your channels, you can build your virtual assistant and all the tasks it
handles once and adjust responses specifically to each channel. This means all development work can
happen upfront, and a virtual assistant can be deployed and then tweaked to differentiate cross-channel
experiences.

CAI bridges the gap between voice and digital support channels using AI, forms, SMS, chat, or email.

While virtual assistants are often called "chatbots," they are not the same. Traditionally, chatbots were only
capable of acting based on strict rules and responding in a matter-of-fact way. Virtual assistants using
conversational AI technology can mimic human-like dialog and give your customers and employees the feel
of a real conversation while helping them with their queries.

A Virtual Assistant Success Story

The bank's new virtual assistant now handles over 80% of the routine queries of their 65 million
customers, with an over 90% call containment rate.

A global financial services company headquartered in New York and operating in over 160 countries was
receiving over 15 million monthly calls regarding partner credit card queries. These credit card partners relied
on the customer service operations of the bank, and call volumes were increasing 10-15% every year. The bank
needed a cost-effective solution that still enabled it to serve customers in the manner they expected.

To do this, the bank launched a virtual assistant built on the Kore.ai Experience Optimization (XO) Platform. The
system was designed to automatically divert customer calls to web and chat channels. Using the advanced AI
and NLP capabilities of the platform, these virtual assistants were able to deliver highly-accurate, human-like
automated responses.

And the result was nothing short of amazing. The bank's new virtual assistant now handles over 80% of the
routine queries of their 65 million customers, with an over 90% call containment rate. It also can interact with
customers in over 25 languages so customers can use their preferred language.

CONVERSATIONAL AI 101 7 2
Re-Defining Experiences with
Virtual Assistants
Virtual assistants optimize both customer and employee experience.
After all, you can't achieve one without the other.

The future of enterprise engagement, productivity, and innovation lies in experience optimization
(XO). In a typical enterprise, the workday consists mainly of interactions between one or more key
stakeholders. These stakeholders include customers, employees, vendors, and the various
systems and software the enterprise uses. But not all of these interactions can be classed as
strategic or "high-value" conversations. In fact, a survey by Kore.ai found that only 25% of
interactions fall into this category. All the rest are routine or repetitive. These routine interactions
are ripe for automation and optimization.

Virtual assistants optimize both customer and employee experience. After all, you can't achieve
one without the other.

Self-service: Customers often know

Optimizing Customer
exactly what they want and would
Experience
rather help themselves than wait on
hold for a human agent.
Using virtual assistants to handle
customer service will create
24/7 access: Customers can ask
personalized customer experiences
questions any time of the day, not just
across both voice and digital
during business hours.
channels. They can respond to and
deflect complex queries with a high
Quick resolution of complex issues:
degree of accuracy, leaving the
Advanced NLP can pick out key terms
high-value tasks to human agents.
in unstructured conversations, store
They not only automate tasks and
this data, and route the customer to the
free up time but optimize the
right channel or agent.
customer experience at the same
time in these ways:
Omni-channel: Conversational AI can
provide the context of the interaction
across every communication channel,
ensuring consistent experiences across
channels.

CONVERSATIONAL AI 101 8 2
Optimizing Employee
Experience
Virtual assistants also optimize
employee experience in many ways.
Because virtual assistants are handling
customer queries, employees can:
But this is not the only use for
them. Virtual assistants can also
be used internally for:
Experience
Focus on rewarding high-value
tasks instead of routine and
repetitive tasks
Answer customer queries that
get routed to them more
quickly witah the contextual
data provided by the
conversational AI system.
CONVERSATIONAL AI 101
Recruitment: HR virtual assistants can
screen potential employees and
trigger background checks.
Ticket Queue: IT teams can deploy
virtual assistants to automate trivial
tasks and assign tickets intelligently.
Sale Report: Easily integrates with
CRM systems and helps sales reps
quickly access reports and
recommendations.
Support: Virtual assistants can answer
questions, provide resolutions to
common queries, and kick off simple
automation.
Vendor Onboarding: Eliminate the
time-consuming paperwork by
deploying virtual assistants to
automate the vendor onboarding
approvals.
9 2
What Can Virtual Assistants Do
for You Today?
Virtual assistants are already helping many industries streamline services. Businesses in
healthcare, insurance, retail, sales, marketing, and more are benefitting.

To see just what today's virtual assistants can do for your business, it helps to see just what they do
for other companies. Here are some ways businesses are using conversational AI to optimize the
customer experience.

Health Care Insurance

In health care, virtual assistants are being
used to diagnose patients and schedule
appointments for them before a doctor is
even involved. These virtual assistants can
smartly capture patient symptoms based on
a chat interaction, suggest a relevant
specialist, book the appointment, and have
all relevant data ready for the doctor on the
day of the appointment. Virtual assistants
are also used in health care to reschedule
appointments, answer questions about
upcoming appointments, and keep track of a
patient's medication list.
In insurance, virtual assistants can
handle accident reports without
involving an agent. The voice or chat
interface of the assistant enables
self-service, guiding users through the
process of reporting an accident. Users
can describe the incident in detail,
provide driver and vehicle information,
and upload pictures of the accident,
streamlining the claims process. They
are also used for making payments,
adding drivers to a policy, and updating
addresses.

Retail Sales and Marketing

Returns and exchanges are more
common in e-commerce because buying
is easy, but making sure you have the
right product is hard when you can't see
a product in person. Virtual assistants
can deliver a hassle-free return
experience for customers without
bothering agents with the tedious
process. They can also help customers
place an order, check an order's status,
cancel an order, and modify existing
orders.
Lead nurturing can be a tedious task.
It involves constant following-up with
prospects, and these follow-ups are
often missed. Virtual assistants can
automate this process with a smart
follow-up sequence and engage
customers on behalf of sales reps.
Virtual assistants can also notify sales
reps when they identify a prospect
that is ripe for a cross-sell or up-sell.

CONVERSATIONAL AI 101 10 2
Conclusion
Virtual assistants aren't the chatbots of the past.

Experience is the total of many interactions, many of them routine. However, as a business
grows, it becomes harder to provide customers with personalized interactions led by a human
agent. In the global race to provide efficient service, every industry you can think of is adopting
conversational AI to make up for this gap.

What they are discovering is that conversational AI and virtual assistants not only improve
efficiency by automating routine interactions, but they also optimize the experience. Virtual
assistants aren't the chatbots of the past. With virtual assistants, enterprises can provide an
effective and scalable way to automate customer inquiries, and customers can express their
needs in natural language and get immediate help.

For details on how conversational AI can improve customer and employee experience, get in
touch with us here.

CONVERSATIONAL AI 101 11 2
About Kore.ai
Kore.ai is a global leader in enterprise conversational AI platform and solutions helping enterprises automate voice and digital
interactions that deliver extraordinary experiences for their customers, contact center agents and employees worldwide. More than
200 global 2000 companies trust Kore.ai’s experience optimization (XO) platform and industry solutions to automate conversations
and deliver extraordinary outcomes for over 100 million users worldwide. Kore.ai is recognized as an enterprise conversational AI
platform market leader by top analysts. The company is headquartered in Orlando, Florida with offices in India, the UK, Japan,
South Korea, Europe, and the Middle East. Visit kore.ai to learn more

Orlando Hyderabad London Tokyo Seoul

www.kore.ai @koredotai www.linkedin.com/company/kore-inc info@kore.com

Copyright © 2022, Kore.ai All rights reserved.

This document is provided for information purposes only and the contents hereof are subject to change without notice. This document is not
warranted to be error free, nor subject to any other warranties or conditions, whether expressed orally or implied in law, including implied warranties
and conditions of merchantability or fitness for a particular purpose. We specifically disclaim any liability with respect to this document and no
contractual obligations are formed either directly or indirectly by this document. This document may not be reproduced or transmitted in any form or
by any means, electronic or mechanical, for any purpose, without our prior written permission.

12 2