A STATE - OF-T H E-MA RKET REPORT

ENTERPRISE
AI TRANSFORMATION

2500 interactions with 1500 enterprise leaders teach

you this: All at-scale problems – across industries –
are of three types. X-ray them further to determine
actionable challenges. Then, design solutions.

15 strategic Challenges are common across industries

and AI can efficiently solve every one of them. Gaming,
grocery or jewelry. Fashion, photography or pharma.
AI is your ally – make no alibi.

From problem analysis to smart fixes, this

State-of-the-market Report on Enterprise AI
Transformation has insights to help you in
your digital transformation journey.
 A STATE - OF-T H E-MA RKET REPORT

ENTERPRISE
AI TRANSFORMATION
Curated with exclusive commentaries and insights
from 29 practicing AI leaders, across industries.
NOTES FROM
LEADERS

ANDY
WALTER

Senior Vice President - IT, Procter & Gamble (Retd.)

 NOTES FROM LEADERS - ANDY WALTER | iv

Andy Walter is a business results-driven professional with

extensive experience in strategy, development, execution,
and operations across Global Shared Services, Analytics, and
IT. He led the Commercial Services & Delivery Organization
(over 1500 IT and multifunctional professionals) for
Procter & Gamble’s Global Business Services (GBS). He
was responsible for IT & Shared Services for all Global
Business Units and Markets around the world. His team was
accountable for developing cutting-edge digital capabilities
for Procter & Gamble to win “where it matters most,”
with consumers, shoppers, and retailers. This included all
eBusiness, Consumer Services, BI/Analytics, Sales Force
Solutions, Project Delivery, Business Process Services, and
A&D / Company restructuring efforts.

He has over 20 years experience working in Boards and in

board-level advisory roles. As a member of the PathFactory
and True Essence Foods Private Board of Directors, he is an
independent director advising on company strategy, cyber-
security, and execution across industries. As a Strategic
Advisor and Investor, he focuses on disruptive innovators in
Digitization, Sustainability, Platform Models, and Advanced
Analytics domains.

He is author of Waiting Is Not an Action – a business

leadership book for executives across industries with
focus on Disruptive Innovation, Strategy, and Career
Development.

His passions include Vintage Racing and Touring classics as

co-CEO (with Laurie, his wife) of L&A Touring Design. He
holds a Bachelor of Science degree in Computer Science
from the University of Cincinnati. Andy and Laurie reside in
Cincinnati, Ohio USA.
THE STATE OF AI -
FOURTH GENERATION
ENTERPRISE
TRANSFORMATION
Andy Walter
Senior Vice President - IT, Procter & Gamble (Retd.)

As a strategic advisor for Mad Street Den (Vue.ai) and

several other disruptive innovators across the technology
and analytics industry, I have seen that we are at a tipping
point for Enterprise Transformation with the acceleration
of AI capabilities. Though scaling over the last two decades,
AI is now at the point of becoming the Fourth Generational
Enterprise Technology disruptor.

Over my 33 years in the industry, I was fortunate to be part

of three previous generational tipping points in enterprise
technology. They all encountered the same challenges of
irrational exuberance, lack of C-Suite vision to leverage
in the short-term, and then subsequent significant value
creation (productivity, business process reinvention, and
totally new-to-the-world business models).

Gen 1 - 1970’s - 1980’s

Wide-scale deployment of mainframes/midrange

computers drive transformation in all parts of the
Enterprise. Slow to adopt at first, the Fortune 500 winners
drove the transformation of every business process,
employee productivity in all corners of the enterprise, and
the birth of completely new industries.
 NOTES FROM LEADERS - ANDY WALTER | vi

Gen 2 - 1985 - 2000’s

Wide-scale micro-computer and mobile technologies

(laptops, smartphones, etc.) created an agile enterprise
with no walls. Massive productivity in back-office,
sales, and overall administration was matched by the
consumerization of technology in every home.

Gen 3 - 1993- 2010’s

Wide scale adoption of internet capabilities internally in

Enterprises and in every aspect of their Go-To-Market changed
everything. Companies winning at scale were now able to be
disrupted by smaller, agile, consumption model-based start-ups.

AI, and more specifically modern Generative AI / LLM birthed in

2019 at Google, will be the Fourth massive transformation / black
swan in my career. Like the first three Gens created incredible
productivity, new business models at established companies, and
entirely new businesses, Gen 4 will drive a similar magnitude, but
at an accelerated pace.

Gen 4 - 2010- 2020+

AI/ML moves from bespoke business process optimization

with LLMs providing a massive range of applications. Gen 4 will
require C-Suite leadership vision (where could we create value),
knowledge (where can we create value now), and execution (how
do we release value). LLMs will disrupt the routine (customer
support/call centers), enhance the complex (application code
development), and transform what is possible (drug discovery). As
with each previous generational transformation, the winners and
losers in each industry will only be separated by the caliber of
their executive leadership.

Play to win and Waiting Is Not An Action!

INSIGHTS
INSIDE
Surfing the AI Wave 01

Notes from leaders 02

Fast-track or fall behind: The Sputnik moment for AI in America
CAROL GRUNBERG, CHIEF BUSINESS OFFICER, YUNO; FORMER MD AT CITI GROUP

From Noise to Signal: A 2024 AI Memo 06

FOUNDERS’ NOTE

Notes from leaders 18

The Human-AI Co-Pilot Model: A New Era of Business Transformation

MURALI SUNDARARAJAN, EVP & CIO, VICTORIA’S SECRET

Dialing, chatting and growing our community 22

What enterprises want 23

Improving Productivity 24

Notes from leaders 26

Why data is and will continue to be the key to a successful AI project
TIMO WEIS, ASSOCIATE VICE PRESIDENT, INFOSYS LTD

Our findings 31

Challenge I | Reduce human errors and improve efficiency 33

Challenge II | Extract, validate and organize data 39

Challenge III | Improve time to go-live 45

Challenge IV | Improve compliance 51

Challenge V | Improve data insights 55

Saving Costs and Time 60

Notes from leaders 62

My predictions for AI in 2024
CHETAN SHAH, SENIOR VICE PRESIDENT - HEAD, PROJECTS & AUTOMATION, INDUSIND BANK

Our findings 66

Challenge I | Decrease dependency on manual resources and processes 67

Challenge II | Find cheaper alternatives for similar solutions 71

Challenge III | Reduce wastage 75

Challenge IV | Reduce discrepancies between teams 79

Elevating Customer Experience 84

Notes from leaders 86

Customer Experience from my lens
ADITYA AGRAWAL, HEAD, CX OPERATIONS, TRANSUNION CIBIL INDIA

Our findings 90

Challenge I | Enable CRO and smoother website journeys 91

Challenge II | Create an outstanding omnichannel experience 97

Challenge III | Craft a seamless onboarding experience 101

Challenge IV | Activate a fantastic Customer Support channel 105

About Vue.ai 110

 SURFING THE AI WAVE | 01

SURFING
THE AI WAVE
The noise around it is deafening.
The complexity? It’s off the charts.

Today, AI might seem like a wild roller

coaster, leaving enterprises with a mix of
confusion and downright fright. But AI is
only a toolbox, not a black box or a cryptic
puzzle even. You are the captain of the ship
and AI is no sea monster. It’s a powerful
compass, though.

We help you cut through the noise and

identify signals, find needles in the haystack
but with a magnet. We strive to make your
learning curve a gentle slope.

We uncomplicate AI.
NOTES FROM
LEADERS

CAROL
GRUNBERG

Chief Business Officer, Yuno;

Former MD, Global Head Partnerships & Innovation, Citi Group
 NOTES FROM LEADERS - CAROL GRUNBERG | 03

With vast experience in the global payments and

commerce industry, Carol is a visionary leader and a
strategic partner who creates the ultimate payments
ecosystem and transforms global commerce. As
the Chief Business Officer at Yuno, a fast-growing
fintech startup, she leverages her expertise in
new business development, client experience, and
strategy to expand Yuno’s reach and impact across
markets and segments.

Carol’s passion for empowering growth and

innovation stems from her diverse and distinguished
career at some of the world’s leading technology
and finance companies, such as Citi, Northwestern
Mutual, Ant Group, and Google. She has been
recognized for her excellence in leadership, product
development, and partnership deals, and has been
invited to speak at prestigious events and forums on
AI, commerce, and payments innovation.

She is also a dedicated investor and mentor for

early-stage startups, especially those focused on
fintech, digital transformation, and social impact.
Carol is committed to driving financial inclusion and
invention for billions of people, and to championing
women and community development.
FAST-TRACK OR FALL
BEHIND: THE SPUTNIK
MOMENT FOR AI IN
AMERICA
Carol Grunberg
Chief Business Officer, Yuno;
Former MD, Global Head Partnerships & Innovation, Citi Group

The year 2017 witnessed a pivotal moment in the rise of AI

capabilities when Google’s DeepMind AlphaGo defeated Ke Jie, the
19-year-old, Chinese world number one Go champion at that time.
While governments all over the world woke up to AI’s newfound
star power, in 2021 and 2022, China established itself as a pioneer in
regulating artificial intelligence by implementing detailed, binding
regulations on common AI applications

The Cyberspace Administration of China’s (CAC) Generative AI

Measures, which came into effect in August 2023, had detailed
mandates on content moderation, data protection and algorithmic
governance, for companies working with AI. AI’s integration in Chinese
companies is not limited to impressive showcases of prowess but
extends deeply into everyday business operations. From customer
service to forecasting and operational efficiency, AI’s footprint in
China spans several sectors, including banking and housekeeping.

Other global economies too are not just experimenting with AI but are
weaving it into the core of their economic strategies, signifying a shift
toward a future where technological superiority could equate to global
leadership. For instance, corporations leverage app download trends
to anticipate consumer behavior across regions, identifying potential
hotspots for monetization. This insight allows early partnerships with
advertisers, securing advantageous positions in emerging markets.
 NOTES FROM LEADERS - CAROL GRUNBERG | 05

Against this backdrop of an era where digital transformation dictates

global leadership, the urgency for America to deepen its investment
in AI cannot be overstated. For America to maintain its competitive
edge, investing in AI is no longer optional; it’s a necessity. This is not
merely a technological ambition but a strategic imperative, especially
as countries like China advance rapidly, integrating AI into the fabric
of their industries and governance.

Financial services, arteries of the global economy, present a compelling use

case for AI’s transformative potential. Here are a few instances:

• Customer Service: By analyzing transaction history, banks are able to

predict the reasons behind customer inquiries with over 90% accuracy.
This pre-emptive approach enhances client experience, satisfaction and
efficiency, showcasing AI’s ability to turn data into actionable insights.

• Cross-border trade: In international finance, AI is able to capture

market anomalies and offer more precise forecasts about global money
movement, helping banks manage foreign exchange fluctuations
and cash liquidity more effectively. By analyzing market conditions,
geopolitical events, and economic indicators that influence currency
values AI helps in managing the risks associated with forex transactions.
This can guide financial institutions and users in setting more precise
hedging strategies to protect against adverse movements in forex rates.

• Predictive modeling & user behavior: Identifying tomorrow’s giants

today is critical in several aspects of the economy, including retail,
venture capital and financial services. AI assists in spotting fast-growing
companies early, enabling organizations to form lucrative partnerships
before these enterprises skyrocket to success. This is not just about
managing cash or equity positions but foreseeing market shifts.

These examples underscore a broader narrative: AI’s potential to

redefine industries is immense and immediate. For America, the race is
not just against time but against a swiftly evolving global backdrop where
technological supremacy is paramount. The country’s commitment to AI will
not only determine its competitive stance but its role in shaping the future
economic landscape as well.
FROM NOISE
TO SIGNAL:
A 2024 AI MEMO
 FROM NOISE TO SIGNAL: A 2024 AI MEMO | 07

Here we are. At the proverbial crossroads!

For the last couple of decades, Anand, Costa, and I have

been on a journey to understand the role of the brain, our
society, systems, people and how they influence the way
we build technology systems. Our conundrums over the
past decade, formulating our thesis for Mad Street Den and
constructing it over the last seven years, have given us a
unique perspective on what has transpired in this time in AI.

If you happen to be the AI evangelist, the active user,

the probable consumer, the decision maker for the
use of AI or the one mandated to “implement AI” in
your organization, this piece is for you!

A Contrarian View
If we were your archetypal Southpark-fiend AI company, this would be
the way that this story would have simply been put:

We are, for better or indeed for worse, not that

company. This is not that piece.
This is our journey of understanding the tight
dependencies of good ol’ software engineering and the
advances in machine learning, resulting in a “what if”
question of an ideal AI-native data platform. Not quite
AGI, but an important step towards that holy grail.

This is our journey of looking around the corner,

anticipating exponential growth in compute power
ubiquity, and fine-tuning network architectures to exploit
that growth, culminating in non-linear effects that are
super exciting, opening doors to new breakthroughs.

AI that glitters is not gold

While the world has been busying itself in going from the
skeptical extreme of will-AI-make-it-out-of-Asimov’s-writing
to today’s near-dystopian how-long-before-sentience, AI
implementations have spawned a generation of people with bad
habits, entire industries applying sub-optimal practices, tools,
processes and applications of these systems. Most importantly,
it has resulted in enormous tech debt that urgently needs
undoing in a rapid manner.

The damage caused is predominantly in three ways:

• Implementation of vanity AI projects, that have essentially

gated meaningful implementation of AI towards meaningful
outcomes.

• Creation of a generation of talent that believes that building

models is the only job worth doing, while paying scant heed
to fixing the data – the lifeblood of all of the models.

• Loss of tens of billions of dollars to sub optimal, fragmented

efforts not driven by business goals.
 FROM NOISE TO SIGNAL: A 2024 AI MEMO | 09

Taming of the (not so) Shrew-d data stack

The journey ahead is about taming the modern data
stack. In the last five years, we’ve successfully created a
multitude of parts that help move, organize, and unify
data. No good deed, though, goes unpunished. AI is no
exception. What has happened is thus:

• Budgets in the clouds: Spiraling cloud costs that

beguile CTOs and Financial Controllers in equal
measure.

• One more thing: A journey from Apple’s ‘There’s an

app for that’ to the hyperscaler’s ‘There’s a modern
data stack component for that’.

• The journey itself is the destination: An architecture

nightmare for data and transformation leaders looking
at an entire enterprise’s data and looking to bring
some sense of progress to it without waiting for a
knock on the door every day from a company that says
“I’ll do this for you” with 20 vendors, 40 moving parts,
a plethora of tools and an interminable rolling five-
year transformation plan.

A Broken Promise
There are two common questions in every enterprise
boardroom today:

• The investor and executive ask: How can we realize the

promise of AI for ourselves? It seems more ‘within reach’
than ever.

• The operator question: Are we realistic about the

readiness, what with our units with divergent processes,
unwieldy data, and talent attrition like never before?
 A two-word phrase, more than
any other, has, in the past decade,
led to billions going down the
drain in its mythical pursuit –
Digital Transformation.

The truth, unfortunately, is that, well, snake oil has many

incarnations, Digital Transformation simply being the
latest. According to MIT Sloan, the cost of failed digital
transformation projects is USD 50B.

The total cost of operations of AI for enterprise is making

leaders everywhere take a long pause. A CEO of a large
company we recently met, told us that they had invested so
much in their multi-year data efforts but attrition in their
organization has outdone these efforts and the last one
standing in their data transformation journey is a third-
party consulting vendor who is not accountable for any of
the architecture decisions made over the last three years.

Digitizing a broken process gets you

a digitized, broken process.
 FROM NOISE TO SIGNAL: A 2024 AI MEMO | 11

The Model Blue Pill

Customer experience does not wait for the

‘transformation project’ train to arrive.

SBUs want results, and they want it now. They unleash their Neo, or
whatever their equivalent of taking the blue pill is. Some models are built
on some data leading to failure of epic proportions, and labeling of AI as a
fraud resulting in increased AI budgets – budgets are funny that way – and
laurels and promotions for transformational leaders. Failing upward has
never been easier.

Bain published a study showing a meager 25% of all models ended up

actually living in production environments in the last few years. Also, the
total cost of building and deploying AI continues to soar; the complexity
of managing disparate systems grows with the talent available — sparse at
best — while time to market and meaningful business outcomes suffer.
Data-Centric + Automation-Centric Approach
to AI: A Promise to Our Customers
At Vue.ai, we have always believed in a data-centric approach
to operationalizing AI at scale.

This has meant a relentless focus on

enriching and extracting the most out of
data, creating taxonomies that live and
change progressively as businesses change,
while letting the AI decide which models get
applied when, for who (and, often, by whom)
based on interaction with its users.
 FROM NOISE TO SIGNAL: A 2024 AI MEMO | 13

This approach has allowed us to apply our platform toward fulfilling

these key promises to enterprises.

• Fix the data in an automated, cost-efficient manner creating a

unified, centralized living multi modal data repository that is
constantly enriching itself.

• Auto-mine data with AI and deploy contextual graphs, rapidly

driving actionable intelligence into applications with real-time,
accurate, relevant information.

• Straight through automation of connected workflows enabling

STP like you have never seen before! (RPA is dead, thanks to the
need for enriched, unified data.)

• Drive actionable intelligence through current enterprise apps

(rapid AI deployment) with proof of financial ROI in 90 days.

Now that organizations are in the early days of a

consolidation cycle of point solutions and multiple
vendors, customers (CFOs and CTOs) need a partner
who fulfils the promise of a simplified architecture,
metrics improvement and who has skin in the game
from raw data to models to apps.
At Vue.ai, we understand this all too well having served large
enterprises over the last 7+ years. With our generalizable AI
engine, we’re ushering in the era of the platform that delivers
outcomes, rather than one that sells tools to a new generation
of model builders.

Vue.ai — Your AI Orchestration Platform

We’ve grown with our customers, as we’ve forayed into retail,
finance, insurance, healthcare, pharma, and more with our AI
suite of products. Across each of them, we realize that all our
customers care about are outcomes, data and automation. A
culmination of this has been a time-tested, outcome-driven,
data-first approach to AI and this has meant a complex
orchestration of three things.

• The efficient movement of clean, multimodal, AI-enriched

datasets that grow and change with time, are connected
to multiple applications with no divergence of data, with
accurate representation of lineage.
 FROM NOISE TO SIGNAL: A 2024 AI MEMO | 15

• Real-time orchestration of models, a federated intelligence engine

applied contextually while observing changing data, interactions
with it from end users, and responding to it appropriately, allowing
for more optimal sets of outcomes.

• Deep connections between workflows, a seamless end-to-end

automated approach for a connected enterprise with minimally-
repetitive human intervention.

AI Orchestration is the ability to seamlessly move between the

3 layers above: enable enriched, alive, progressively-changing
data sets for entire organizations and across teams; enable use
of a federated engine that contextually understands what the
user wants and responds with the right models; and a smooth
coordination of end points across workflows creating smooth
journeys across your enterprise.

For example, if you’re looking to enable

personalization, we layer intelligence across search,
recommendations, cross-sells and upsells, forecasting,
notifications, support, customer interactions from
everywhere. If you’re looking to enable document
automation, we layer intelligence across extraction,
classification, matching, reconciliation, feeding all
of the derived intelligence into unified customer
profiles and centralized data systems for everyone,
feeding right back to customer experiences or optimal
workflow management in your organization.
With this renewed focus on orchestration, we
promise to help your entire organization adopt AI
across its functions, undoing the messiness of the
last 5+ years.

We believe orchestration simplifies.

Here to serve your analytics, IT & business

teams as their System of Intelligence
We’re at a time where AI tools and solutions are almost a
commodity. The question for the enterprise today is not
one of which applications to buy or build. The steep path
ahead is one of orchestration across its functions, different
departments, each of their needs, and the state of each of
their data.

At Vue.ai, we have a deep, long-standing partnership with

large enterprises and partners around the globe who
put their data’s quality and their vision for a unified and
connected system before a large budget for model building.
We have successfully become the System of Intelligence
and, increasingly, also becoming a System of Record for our
trusted customers who will constantly mine meaning and
context from all that data, keeping it readily accessible by
applications from IT to business units.

Our suite of native applications, Vue.ai Hotstart, along with

our preset, industry-specific model library have made for
the perfect entry point for business teams to begin their AI
journey towards an outcome-centric AI adoption.
 FROM NOISE TO SIGNAL: A 2024 AI MEMO | 17

Whether it’s personalization,

document processing and process
automation, content tagging or
forecasting and predictive analytics,
our native applications focus on
providing quick resolution.

We expose our model-building tools and our

data stack to our customers as well, allowing
for them to work off of a single orchestration
platform to mediate data, and to serve all their
AI and automation needs.

Towards that vision, at Vue.ai, we uncomplicate.

(We were thinking of coining something to the
effect of complify or simplicate but that would
not be uncomplicating, would it?)

WELCOME TO VUE.
AI, UNCOMPLICATED.
NOTES FROM
LEADERS

MURALI
SUNDARARAJAN

EVP & CIO, Victoria’s Secret

 NOTES FROM LEADERS - MURALI SUNDARARAJAN | 19

Murali is a global Fortune 500 executive leader and a

Board member with strong general management and
operational leadership experience in the retail industry
for the last 30+ years. He has extensive experience in
leading a global business unit while serving as a Managing
Director in Asia, and currently overseeing IT globally as
EVP Chief Information Officer. He is a proven, results-
oriented technology innovator (patent owner) with extensive
knowledge and experience overseeing domestic and global
technology operations in various markets, including Asia,
UK/Ireland, and Europe. Murali is widely respected as
a highly engaging and intuitive leader with exceptional
relationship-building and business development skills.
THE HUMAN-AI
CO-PILOT MODEL:
A NEW ERA OF BUSINESS
TRANSFORMATION
Murali Sundararajan,
EVP & CIO, Victoria’s Secret

What was once considered a business expense

– technology and its associated infrastructure
– is now an indispensable asset for achieving
enterprise success. Advances such as Hadoop,
IoT, and robotic process automation have been
instrumental in this transformational change,
streamlining processes and unlocking the value
of data. The Covid-19 pandemic underscored the
importance of convenience and immediacy in
service delivery, necessitating rapid data curations
and real-time pattern recognitions. It has shifted
our models from being on autopilot to becoming
co-pilots that can react and adapt swiftly.

In my interactions with various customer service

channels, I see a growing preference for digital
experiences and immediate solutions. Customers’
expectations too are steadily evolving towards
instant gratification. From a financial standpoint,
businesses should strive for optimal efficiency to
liberate resources for growth and transformation.
 NOTES FROM LEADERS - MURALI SUNDARARAJAN | 21

Automation and Artificial Intelligence are integral to

this pursuit. While the need for clean data is critical,
AI is only a layer that sits and operates on top of this
clean pool of data. Large Language Models such as
GPT hold the potential to revolutionize industries
by significantly reducing the time spent on data
curation and mapping. For instance, automating the
FDA medical approval process with AI could reduce a
procedure that formerly took five months to a mere
five minutes.

Navigating the intersection of technology, data,

and human adaptation requires a keen focus on
managing change and driving efficiency, without
any compromise on accuracy. A culture of proactive
innovation built on the principles of continuous
quality improvement is vital to expand the frontiers
of our businesses. In this era of data-driven
decision making, the adage “garbage in, garbage
out” holds true more than ever. While technological
advancements have surged ahead, processes and
people are still catching up. A paradigm shift is
needed: humans should focus on validating data and
trusting the outcomes generated by AI, rather than
attempting to replicate machine tasks.

I find it easy to understand and explain this changing

paradigm using the maze as an analogy. It is
extremely important for everyone to understand the
relative position of each individual in the maze and to
guide each other towards their respective goals with
flexibility. Successful individuals can serve as anchor
points to inspire and motivate others.

As we navigate this changing landscape, it becomes

apparent that adaptation remains one of the greatest
challenges. To truly realize the full potential of
AI-powered efficiency, we must ensure that our
technology, processes, and people are aligned in
terms of speed and efficiency.
DIALING, CHATTING,
AND GROWING OUR
COMMUNITY

How this report was compiled

Over the past 12 months, we have dialed, swiped, and
tapped our way through some 2500 calls about AI,
digital transformation, and everything in between. About
500 enterprise CxOs and VPs are already a part of our
events, podcasts, roundtables, and more. And another
500 of them are not just listeners; they’re our allies in
intelligence, sharing their love for all things AI. Our AI
transformation retreat REBUILD is a family reunion but
with more algorithms and fewer awkward hugs.

Almost a 1000 of our conversations this year are swirling

in the enchanting vortex of the buying cycle. Improving
business productivity, saving costs and time, and
elevating customer experience were the key touchpoints
of our discussions. From extracting, organizing, and
validating data and mining them for insights, to reducing
dependencies on manual resources and processes to
delivering stellar omnichannel experiences and CRO,
nothing brings us greater joy than setting up AI and
automation that can step up and deliver.

Here’s to another year of conversations, innovations,

executions, and maybe a few marquee events thrown in
for good measure.

Stay tuned!
 WHAT ENTERPRISES WANT | 23

WHAT ENTERPRISES
WANT

Across 10 industries and 125 enterprises, our

conversations with business and technology leaders
over the past 12 months informed their pain points
and signaled their unmet needs. Three broad, thematic
opportunities capture their challenges.

a) Improving productivity

b) Saving costs and time

c) Elevating customer experience

How can AI help alleviate these challenges?

Where exactly are the opportunities for improvement,
in each of these themes?
 IMPROVING
PRODUCTIVITY
 IMPROVING PRODUCTIVITY | 25

An annual addition of USD 15T to the global economy is the

expected productivity gains, informs a McKinsey Global
Institute study on deploying AI. But only after overcoming
numerous technological, process-related, and organizational
hurdles and industry dynamics that stand in the way.

For the past three decades, backed by the exuberant

growth of productive capacity in China and other emerging
economies, which kept inflation in check, central banks
kept interest rates near zero. As all good things come to
an end, many developed economies are now witnessing a
slowing of growth and productivity. In the United States,
for example, labor productivity growth which was at 1.73%
in the decade before the financial crisis dropped to 0.53%
in the decade before the pandemic. On the supply side, in
countries that account for more than 75% of global economic
output, dependency ratios – the number of non-workers
relative to the working-age population – are rising due to
aging populations that have limited the growth of the labor
supply and are creating fiscal stress. Many large employment
sectors such as the government, health care, traditional
retail, hospitality, and construction face critical shortages
of workers. And overall labor forces are shrinking in some
countries, like China, Italy, Japan, and South Korea.

Challenged by the pandemic, geopolitical tensions,

and climate change, companies and countries are
very motivated to “de-risk” and diversify their
supply chains at great expense, for reasons that
have nothing to do with reducing costs.

Against this backdrop, AI holds the potential for a digital-enabled

surge in productivity that could restore growth momentum by
enabling value-added growth, in which firms and sectors expand
value-added output, thereby contributing to a rise in GDP, rather
than simply by reducing inputs, such as labor.
NOTES FROM
LEADERS

TIMO
WEIS

Global Head of Growth, Infosys Ltd

 NOTES FROM LEADERS - TIMO WEIS | 27

Timo Weis is a seasoned executive and Global Head of

Digital Strategic Growth at Infosys Ltd. With a proven track
record in driving revenue and transforming businesses into
digital powerhouses, Timo is a recognized expert in digital
marketing and strategic growth.

Throughout his career, Timo has successfully led digital

transformation initiatives in the CPG and Retail industries.
He has a strong background in developing and executing
strategies that engage the right audience and drive results.

Timo’s leadership prowess extends to building high-

performing teams, leveraging data-driven insights, and
staying ahead of industry trends. His professional journey
includes senior leadership roles at renowned organizations
such as Akamai Technologies and Adobe.
WHY DATA IS AND WILL
CONTINUE TO BE THE
KEY TO A SUCCESSFUL
AI PROJECT
Timo Weis,
Global Head of Growth, Infosys Ltd

Most enterprise companies face several data

challenges when implementing artificial
intelligence, including the following:
1. Data Quality and Quantity: Ensuring the data is accurate,
clean, and sufficient to train reliable AI models.

2. Data Integration: Combining data from various sources and

formats into a unified system that AI can analyze effectively.

3. Data Privacy and Security: Protecting sensitive information

from breaches and ensuring compliance with data
protection regulations.

4. Bias and Fairness: Mitigating biases in AI algorithms that

can lead to unfair or unethical outcomes.

5. Data Management: Effectively storing, processing, and

retrieving large volumes of data for AI applications.

6. Talent and Expertise: Finding and retaining skilled

professionals who can manage and interpret complex AI
systems.

7. Scalability: Scaling AI solutions across different departments

or business units within an enterprise.

8. Interpretability and Explainability: Making AI decisions

transparent and understandable to stakeholders.
 NOTES FROM LEADERS - TIMO WEIS | 29

These challenges require careful planning,

robust data management strategies, and ongoing
monitoring to ensure AI systems are effective,
ethical, and compliant with regulations.

Data is the fuel that powers artificial intelligence,

as it enables the training, testing, and validation
of AI models, as well as the generation, analysis,
and interpretation of AI outputs.

Data is also the source of value and impact

that AI can create for various domains and
sectors, as it can reveal insights, patterns, and
predictions that can inform decision-making
and action.

Therefore, data is the most critical factor

that determines the success and failure of AI
projects, and it creates significant challenges
at both the strategic and tactical levels.

At the strategic level, data requires a clear and

coherent vision that aligns with the objectives
and expectations of the AI project, as well as
the needs and preferences of the stakeholders
involved.

This includes defining the data requirements

and specifications, such as the type, format,
size, quality, and availability of the data, as
well as the data sources and methods, such
as the collection, acquisition, integration, and
preprocessing of the data. Moreover, data
requires an ethical and legal framework that
ensures the protection and respect of the
rights and interests of the data owners and
users, such as the privacy, security, consent,
and ownership of the data, as well as the
fairness, transparency, and accountability of
the data processing and usage.
At the tactical level, data requires a robust
and reliable infrastructure that supports
the data management and governance,
such as the storage, access, distribution,
and backup of the data, as well as the data
quality and integrity, such as the verification,
validation, cleaning, and enhancement of the
data. Furthermore, data requires a skilled
and competent workforce that can manage
the data operations and analytics, such as
the extraction, transformation, loading,
and visualization of the data, as well as the
data science and engineering, such as the
modeling, optimization, evaluation, and
deployment of the data-driven AI solutions.

In conclusion, data is the foundation and the

biggest challenge for most AI projects, and
it demands a comprehensive and balanced
approach and action. By addressing the data-
related challenges, businesses can ensure
that they have the right data to get the right
answers from AI, and that they are prepared
to use AI effectively and responsibly.
 OUR FINDINGS | 31

OUR FINDINGS
 CHALLENGE I - REDUCE HUMAN ERRORS & IMPROVE EFFICIENCY | 33

CHALLENGE - I

REDUCE
HUMAN ERRORS
& IMPROVE
EFFICIENCY
A 99.9% accuracy in detecting network
intrusions or even a 98% accuracy in
classifying malignant email attacks are
the kind of numbers that would make
the C-Suite sleep well. All the more so if
it’s highly reliable, accurate and always-
on in domains where human intuition
generally fails. Strides made in machine,
deep, reinforcement, and NLP learning
methodologies are phenomenally
accurate in improving business
efficiencies by reducing opportunities
for human errors. Our conversations
with business practitioners unveiled
many common scale-and-speed pain
points that could be solved using AI.
 CHALLENGE 1 - REDUCE HUMAN ERRORS & IMPROVE EFFICIENCY | 35

Stumble Street: Where enterprises hit

the speed bump
1. Standardizing the error-prone metadata
generated by in-house and outsourced teams
at online marketplaces, leads to slower go-live
to the market and the lost opportunity cost of
realizing revenues sooner.

2. Hustling a team just in time to create product

data. At off-price stores especially, there is
little visibility on the items they receive from
manufacturers, till they receive them.

3. Extracting information from vehicle

registration cards and reconciling it with
the data provided by users at used-car
marketplaces. Censoring NSFW keywords,
phone numbers, and unrealistic promises in
vehicle descriptions. Detecting scratches and
dents inside and outside the vehicle, while
comparing image similarity and text similarity
to detect duplicate entries.

4. Induction of caterers by food delivery partners;

validation of mandatory documents - Licenses,
cuisine tags, menu cards, bank details,
contracts, MIF, CR; checking compliance of
product descriptions and imagery to stated
guidelines; scraping competitors’ sites to check
pricing comparison

5. Compliance with regulatory market

restrictions in controlled geographies also
necessitates intense human intervention, a
source of errors and inefficiency.
Smart fix:
Allow teams to focus on higher-value tasks by
accelerating workflows. Improve productivity
by automating manual tasks. Vue.ai’s Intelligent
Document Processing (IDP) and Automated
Product Tagging tools help in handling data
efficiently and creating active learning loops.

• Minimizes manual intervention and reduces the

after effects of human fatigue. The streamlined
document processing improves efficiency and
ensures error-free workflow.

• Uses advanced OCR & NLP technologies built

on Computer Vision with HITL capabilities, to
ensure no errors.

• Reduces errors by extracting and predicting

data accurately.
CHALLENGE 1 - REDUCE HUMAN ERRORS & IMPROVE EFFICIENCY | 37

The homestretch:
Digitization leads to better product
discovery and improved productivity
in such scenarios. For example, a
luxury fashion retailer in the USA that
used to manually manage its product
catalog improved efficiency to tag
a batch – from 5 days to 3 days – by
deploying Vue.ai.
 CHALLENGE II - EXTRACT, VALIDATE AND ORGANIZE DATA | 39

CHALLENGE - II

EXTRACT,
VALIDATE AND
ORGANIZE DATA
When even the most rudimentary, AI-based
extraction techniques are implemented,
organizations require 40% fewer hours
to process routine paperwork, PwC has
disclosed. Validation of the extracted data
based on either rules or images, a common
downstream application, is turbocharged
with ML, NLP, and AI overall. Massive gains
are realized with the use of AI to leverage
raw data – that was the common motif
of our discussions with customers and
associates over the past year.
 CHALLENGE II - EXTRACT, VALIDATE AND ORGANIZE DATA | 41

Stumble Street: Where enterprises hit

the speed bump
1. Communications and counteroffers between
buyers and sellers, negotiations, and reviews
before purchase orders are finally placed
involve heavy manual work. Automating data
extraction from documents can ease the pain.

2. Merging data extracted from email

communications and product metadata with
information sourced from, say, a retailer’s
Purchase Orders or a container shipment
sheet; reconciling the extracted information
with the details already present in a CRM;
structuring this knowledge to match the
vendor’s specific Order Export file format
based on each Purchase Order.

3. Loan application documents are typically

expansive, requiring plenty of data to
be extracted (payslip, credit history etc)
and validated (educational certificates).
Onboarding BFSI customers through KYC
documents, onboarding staffers in industries
according to specified prerequisites,
digitizing the entire healthcare records
of patients in hospitals and mining their
diagnostic test reports for critical alerts are
all situations where AI could help heavy lift
the data processes.

4. When sellers on online marketplaces do not

provide all the required information or even
the right information, when house guidelines
for PDPs are flouted and subverted, strong
data validation methodologies save the day.

5. The unstructured information is available

in dynamic tables whose structure changes
with every document are a unique class of
challenge. NER models can be used to extract
data and populate the CRM.
Smart fix:
Automated Product Tagging improves the
productivity of data extraction by:

• OCR and image recognition systems extracting

data from unstructured or semi-structured
text or images. This extraction encompasses
various data forms, such as text embedded in
images. For instance, the systems can extract
ingredients from labels for grocery products.

Tagging also enhances productivity

through data organization by:
• Standardizing and organizing the chunk of
information received from suppliers and
vendors. In other words - take all the broken
pieces and put them together once and for all.

Data Validation through Tagging

ensures:
• The data in the catalog undergoes tagging,
screening, and flagging for rejection reasons in
a single iteration. This process provides richer
feedback for sellers, contributing to better
quality over time. Vendors can be notified
initially if/when they publish data that does
not match the business guidelines.
 CHALLENGE II - EXTRACT, VALIDATE AND ORGANIZE DATA | 43

The homestretch :
Template-based extraction modes typically
require time and training to solve these
challenges. As demonstrated with a container
shipping major, Vue.ai’s algorithms adapt to
diverse document formats and structures and
are trained to perform without training.
 CHALLENGE III - IMPROVE TIME TO GO-LIVE | 45

CHALLENGE - III

IMPROVE TIME
TO GO-LIVE
Go-to-market teams create full accountability
for increased win rates, shortened sales cycles
and improved forecast accuracies, using AI and
automation. Integrations with different systems
such as marketing automation, CRM, email,
calendaring, telephony, content management
and conversations capture all contact and
activity. Combining execution history with
real-time activity drives the product pipeline,
improves conversions and reduces churn. In our
discussions with AI transformation professionals
charged with improving the time to go-live, we
offered solutions for a variety of challenges.
 CHALLENGE III - IMPROVE TIME TO GO-LIVE | 47

Stumble Street: Where enterprises hit

the speed bump
1. Multi-brand sellers receiving goods from
varied suppliers need to efficiently standardize
all product data across vendors and quickly go
to market as a unified platform of offerings.
Likewise, generating product attributes
based on image tagging was a surefire way to
accelerate rollouts vs the traditional, manual
tagging process at another Netherlands-based
retailer.

2. Marketplaces with AI-generated product

metadata make money from Day 1 compared to
the not-so-uncommon 14-21-day delays that
their manual-tagging compatriots suffer.

3. Using Vue.ai’s deployment, a luxury fashion

retailer in the USA has more than tripled
its manual, per-day throughput in catalog
management: Compare the GTM velocities of
60 images per person, per day manual output
with the 200 SKU’s via automation!

4. As much as online marketplaces and sellers

mandate standardized product details,
descriptions and images, merchants often
circumvent in creative ways, unwittingly
delaying the GTM. Automatic correction
mechanisms with some human supervision
becomes the faster route to go-live.

5. Imagine moderators at a used car marketplace

looking at 1000 to 2000 listings manually per
day – each with 21 images! Excess inventories
of perishable goods bring in another layer of
complexity with their shelf lives.

6. Operators in consumer loan processing and

insurance businesses face similar expectations
of short turnaround times in decision making
without compromising on the integrity of
processes.
Smart fix:
Vue.ai’s IDP and Automated Tagging improve
time to go-live by:

• Significantly speeding up onboarding

processes. Vue.ai’s IDP accelerates data
extraction, verification, and overall document
handling, contributing to a faster time to
go-live for new users, clients, or projects.

• Ending the numerous vendor calls, seeking

iteration, and automating the process. This
helps the sellers push their products live
faster and sell products quicker and at the
right time.
 CHALLENGE III - IMPROVE TIME TO GO-LIVE | 49

The homestretch:
Building a knowledge graph of 2.5
billion molecular entities, 200 million
compounds and 20 million chemical
relationships is no mean task. But
once it did that in collaboration with
CAS, Dow Chemical is making giant
strides across its plastics, silicone,
polyurethane businesses, as well
as for many industrial solutions.
Identifying eight to 10 promising
molecules to yield better paper pulp
for one of its projects, using AI, Dow
whittled it further down to four most
promising candidates, one of which it
expects to commercialize next year.
 CHALLENGE IV - IMPROVE COMPLIANCE | 51

CHALLENGE - IV

IMPROVE
COMPLIANCE
In Japan, financial operators came
together on a shared, PoC platform
to pool together previously filed
suspicious transactions. Screening and
monitoring using AI then successfully
assisted in triaging transactions against
sanctions lists and flagging potentially
fraudulent operations.
 CHALLENGE IV - IMPROVE COMPLIANCE | 53

Stumble Street: Where enterprises

hit the speed bump
In our own interactions with businesses,
adherence to internal and external,
regulatory requirements is a critical
component of enterprise operations across
geographies worldwide. Beginning with the
differential computation of taxes for the
same product across different states within
the same country, to compliance with federal
rules of disallowing transactions of certain
prohibited items to business-mandated
controlled release of product imagery, AI
simplifies and facilitates faster adherence to
the stated guardrails.

Smart fix:
Intelligent Document Processing solution
improves compliance by

• Fortifying control over information in

data operations and enhancing data flow
integrity.

• Automating cross-verification against

predefined criteria by consistently
applying standards and rules, mitigating
the potential for regulatory issues.

• Prioritizing the confidentiality and

integrity of customer data. Vue.ai adheres
to stringent compliance standards,
including ISO, AICPA SOC 2, and GDPR.
The homestretch:
Where existing methods of examination
return false positives upwards of 95%,
regulators, law enforcement, and financial
industry professionals now routinely use
AI to monitor and curb financial crime,
screen transactions against Anti-Money
Laundering and sanctions lists, and, flag
high-risk, anomalous exchanges, for
greater regulatory compliance.
 CHALLENGE V - IMPROVE DATA INSIGHTS | 55

CHALLENGE - V

IMPROVE
DATA INSIGHTS
Automobiles with 50+ sensors collecting data
on speed, emissions, fuel consumption, use of
resources, and security. Beverage companies
discerning novel, personalized product
opportunities by scrutinizing as many as 14
different data types across 28,000 data sources.
Clinicians hand-picking their choice of AI and deep
learning methodologies to develop sophisticated
and specialized approaches towards diagnosing
oral cancer with high sensitivity, specificity and
accuracy. Across the entire spectrum, enterprises
are submitting their data troves to intelligent
engines and gaining visibility around the corner.

Conversations over the past year with our network

of industry leaders unearthed similar needs
amongst diverse industries.
 CHALLENGE V - IMPROVE DATA INSIGHTS | 57

Stumble Street: Where enterprises hit

the speed bump
1. Querying legacy documents (contracts,
certificates of incorporation etc) for validity
and expiry, to generate auto-alerts and to
answer questions.

2. Digitization of unstructured and antiquated

judicial records such as stenographer’s notes,
evidence documents and court rulings

3. Incorporating geospatial tagging to lead score

risk assessments for, say, a certain topology’s
propensity towards flooding, crime etc

4. Tightening up supply chains by tracking

people, predicting customer churn,
forecasting demand and developing
competitor intelligence

5. Inventory optimization, transport label

recommendation, pricing recommendation,
fulfillment plan generation
Smart fix:
Data insights are improved by:

• Enhancing data accuracy and consistency,

providing reliable information for analysis,
and leveraging predictive analytics for
elevated insights.

• Providing the ability to derive valuable

insights. For instance, an eCommerce
company with enriched data can analyze
its inventory to determine which tags
perform well each season.
 CHALLENGE V - IMPROVE DATA INSIGHTS | 59

The homestretch:

Chef:
This is our best menu item.
It sells more than any other.
AI:
Oh, well, if it really is your best
menu item, how come repeat
customers of your restaurant
don’t order it again?

Data and analytics dissect consumer

behaviors and even trail them from one store
to the next. Through tokenization, they can
well see a returning visitor. What beverages
do they prefer? Are they cheese lovers?
Conservative spenders? Through their credit
card footprints, diners’ tastes can be followed
to other restaurants too.

The viewer ratings that Netflix uses to

recommend a movie for you to watch amount
to only 2% of real data; the rest is just
extrapolated. ‘Collaborative filtering’ is the
name of this algorithm that the pharmaceutical
marketing industry has now borrowed, to
formulate routine communications messages
with healthcare professionals. From using
GAN’s to discover new molecular candidates
against cancer to creating financial wellness
programs for clients based on spending habits,
enterprises all over are harvesting profitable
insights from their everyday operations.
 SAVING
COSTS AND
TIME
 SAVING COSTS AND TIME | 61

Probabilistic modeling and decision

engines have been designed to counter
annual losses of USD 3B - 5B in the
solar power industry, due to ‘soiling’. By
incorporating historical weather patterns,
particulate matter and other sources
of relevant data into machine learning
algorithms, soiling loss can now be
predicted at any location in the USA and
to understand regions that are best suited
for development with minimal impacts.
Similarly, when the average cost to service
a non-performing loan is 12 times that
of a performing one, automating the
loss mitigation process to make faster
and more accurate decisions about loan
modifications and other workouts makes
eminent sense. Mortgage companies and
banks have a growing interest in reducing
the high cost of distressed-loan servicing,
minimizing the time during which their
servicing units have to advance funds
while borrowers aren’t paying.
NOTES FROM
LEADERS

CHETAN
SHAH

Senior Vice President - Head, Projects and Automation, IndusInd Bank

 NOTES FROM LEADERS - CHETAN SHAH | 63

Chetan heads the Projects and Automation

team for the Operations vertical at IndusInd
Bank. He was responsible for the first end-
to-end RPA initiative of the bank and has
implemented key transformational projects.

He has also spent about 18 years with Citibank

in India and US, across Trade, Wealth
Management, and Treasury and at India Retail
bank as head of digital transformation.
MY PREDICTIONS
FOR AI IN 2024
Chetan Shah,
Senior Vice President - Head, Projects and Automation, IndusInd Bank

Well post all the hype and hoopla over the AGI led by Open AI,
things are coming back to normal, the new normal. The new
normal is that people have realized that AI is here to stay. So
one open question that we had in the past, whether AI is a fad
or not, is answered. As individuals or as corporations we have
to decide how we live with it and how we leverage it for our
individual benefit and derive moats for our organization.

My personal predictions for this year are …

1. Most organizations will have to embark on AI journey in
one way or form. The laggards or disbelievers of past years
will have no option but to turnaround.

2. Organizations that do not, will see ramifications in terms

of not only business, but the willingness of people joining
them, and you are only as good as your team. As such they
would be squeezed from all directions.

3. There is going to be a major change in focus as regards the

focus on requirement of collecting good data. Until now,
the focus on data came as a next step in the AI journey.
Now, the data will be the first step as organizations start
asking the question ‘Do we have the data?’

4. Movement away from ROI being the parameter of decision

making. The focus will shift to being future-ready.
 NOTES FROM LEADERS - CHETAN SHAH | 65

5. Training by organizations to make individuals AI-

friendly. This can have a significant impact on GDP
with the same capital allocation.

6. Expectations from senior management will be real.

AI is an enabler and will not replace the existing
teams / people 100%. It’s not a one-to-one game.

7. With AI becoming pervasive in work life, the anxiety

amongst Gen X is going to increase, leading to
increased health issues. Look out for this!

8. Gen Z and millennials with more time on hand

will- start focussing and nurturing their interest
and experiential activities, leading to a happier and
satisfied generation vs Gen X.
OUR FINDINGS
 CHALLENGE I - DECREASE DEPENDENCY ON MANUAL RESOURCES & PROCESSES | 67

CHALLENGE - I

DECREASE
DEPENDENCY
ON MANUAL
RESOURCES &
PROCESSES
Time-consuming and error-prone coding of
invoices used to require manual comparisons
of previous transactions and input from
individuals across the organization. With AI
reducing the dependence on manual resources
by predicting the correct coding on an invoice
and purchase order, businesses get 40% more
man hours to focus more time on value-driving
projects, besides getting spend visibility.
Queries, emails, messages from customer
portals etc are automatically tagged without
human intervention, enabling ticket deflection
and instant replies to commonly raised issues.

Businesses discussed similar challenges with

our leaders in addressing their dependency on
manual resources.
CHALLENGE I - DECREASE DEPENDENCY ON MANUAL RESOURCES & PROCESSES | 69

Stumble Street: Where enterprises

hit the speed bump
1. An outdoor gear brand and a sportswear
brand that shared similar challenges –
automating the photoshoot processes
of products on human models.

2. A 3D model rendering company that

requires rapid, large-scale deployment
of on-model imagery of products.

3. A studio that does manual photoshoots

of on-mannequin images of garments
for various companies.

4. A fashion marketplace for clothing,

jewelry, beauty and shoes and other
fashion retailers evaluating options to
automatically tag their products.
Smart fix:
Vue.ai’s solutions like IDP and Automated
Product Tagging ensures businesses save costs
across multiple touchpoints, without sacrificing
productivity.

The two solutions work to decrease dependency

on manual resources & processes by:

• Accelerating document processing workflow

with a comprehensive selection of in-house,
pre-trained ML models designed to handle
various document types.

• Automating product onboarding and ensuring

flagging and feedback for incomplete data,
inaccuracies, duplicated information, and
content that deviates from established
business standards.
CHALLENGE I - DECREASE DEPENDENCY ON MANUAL RESOURCES & PROCESSES | 71

The homestretch:
Despite having rules-based systems and
manual audits, missed charges and failure to
meet the charge-capture potential are well-
established detriments across healthcare
setups. Given the climbing rates of claims and
prior authorization complexity and requests,
revenue cycle departments in hospitals
now route the insurer’s answers to assigned
systems and accounts needing remediation
to specialized staff.

No more armies of staff to communicate

with payors about patient cost estimates,
authorization requests and claim status.
 CHALLENGE II - FIND CHEAPER ALTERNATIVES FOR SIMILAR SOLUTIONS | 71

CHALLENGE - II

FIND CHEAPER
ALTERNATIVES
FOR SIMILAR
SOLUTIONS
Why did the weatherman go
back to school?
Because he wanted to become
‘brighter’.

Juggling pressure, temperature, moisture

and a multitude of such dynamically-changing
fronts to predict the weather accurately
has been a tough call. Luckily, AI is excelling
in doing them all and how! A 1,000 times
cheaper, 10,000 times faster. Drug discovery,
paralegal work, healthcare claims processing,
invoice processing and accounts payable
functionalities are among the other leading
domains that use AI to deliver cheaper,
alternative solutions to business problems.

Stumble Street: Where enterprises

hit the speed bump
1. A retailer deploying disparate point
solutions for product tagging and
personalization of goods.

2. A brick-and-mortar department store with

a growing online business, looking for a
cost-effective alternative to replace their
legacy systems.

3. Finding human models willing to pose for

discounted products and their costs per
SKU are common confounders.

4. Inability of the AI partner to scale their

solutions to international markets, leading
to cost escalation.
 CHALLENGE II - FIND CHEAPER ALTERNATIVES FOR SIMILAR SOLUTIONS | 73

Smart fix:
Vue.ai provides a cheaper alternative by ensuring:

1. A cost-effective go-to tool for manual data

entry or traditional document processing
methods. Automation leads to faster and more
accurate results, reducing the overall cost per
transaction.

2. Minimal Turnaround Time (TAT) and

expenditures to enhance the proficiency of
skilled workers, unlocking heightened business
value.

3. Affordable scaling that adapts to your expansion

without a significant cost increase and less
stagnation of resources by smarter allocation.
The homestretch:
From the board to your peers, everyone
wants quick outcomes. Vue.ai’s 30:60:90
plan puts you on the course for your first,
quick AI ROI win. Get use case-to-value
demonstration in 30 days from data access.
A viable plan to realize ROI comes in the
next 30. Within three months, we promise
clear visibility on an accelerated growth
plan of implementation.
 CHALLENGE III - SAVE COSTS BY REDUCING WASTAGE | 75

CHALLENGE - III

SAVE COSTS
BY REDUCING
WASTAGE
The manufacture of advanced composite
parts for the aerospace and defense sectors
is an area where AI has shown significant
reductions in material wastage. Besides
spotting opportunities to use up material
remnants and short rolls that optimize
material yield, AI reduces the length of
cut plans by automating their creation,
creating optimal dynamic plans based on
on-hand inventory and incoming work
orders to maximize material utilization. In
the retail industry – fresh foods at groceries
especially – predicting demand for SKU’s
at a location level is quite impossible to do
with just spreadsheets.
 CHALLENGE III - SAVE COSTS BY REDUCING WASTAGE | 77

Stumble Street: Where enterprises hit

the speed bump
1. An ambitious venture to dispatch a tailor to
customers’ homes, along with a variety of
fabrics and garment designs to pick from.

2. Preview and visualize 3D-rendered clothing

designs on models, before green flagging
them for production.

3. Going one up on the 3D-rendered clothing

designs, an outerwear major required
tagging ‘sketches’ of its products – not even
‘images’.

4. Perishable goods with limited shelf life need

optimal distribution and sale strategies in
order to reduce wastage.

Smart fix:
Vue.ai solutions work from the first touch point
to ensure reduced wastage and help save costs.

• IDP’s efficiency in data extraction and

verification minimizes such wastage.

• The reduction in errors and rework

contributes to cost savings by preventing
the need for additional resources to rectify
mistakes, ensuring a more streamlined and
cost-effective operation.

• Vue.ai’s solution also sidesteps essential data

generation tools or manual data extraction
and standardization processes, which can
significantly waste time, effort, money, and
resources.
The homestretch:
Some 300 target molecules usable in
agrochemical and pharmaceutical industries
can be made from waste compounds produced
at commercial scales: substrates that would
otherwise have been stored, disposed of or posed
environmental hazards. Combining computational
synthesis and AI algorithms has accelerated the
discovery routes of such usable target molecules
from a finite set of byproduct-turned-reactants.
Industrially-realistic, miniaturized, automated
suite of pharmaceutical manufacturing systems
and flow chemistry platforms validated the
viability of the synthetic routes generated by such
wastage-to-valuables algorithms.
 CHALLENGE IV - REDUCE DISCREPANCIES BETWEEN TEAMS | 79

CHALLENGE - IV

REDUCE
DISCREPANCIES
BETWEEN
TEAMS
Outdated legacy applications often operate
as silos in vendor-proprietary formats. Data
silos tend to expand, but not in significance,
a considerable portion succumbing to
redundancy, obsolescence and triviality
(ROT). ROT data, along with its backups,
occupies valuable space with no ROI to the
organization. Effective teamwork becomes
elusive then, when participants lack
uniform access to information.

As islands of information lie scattered

across an organization, knowledge remains
untapped for value, insight, and business
advantage. Bereft of universal sharing and
updates, siloed information gives rise to
inconsistencies, inaccuracies, and integrity
issues. This renders the data unusable for
analytics applications and fails to meet
the operational needs of the organization,
especially as data moves, copies and
version changes across diverse locations.
The most significant concern lies in the
vulnerability of information housed in data
silos, encompassing endpoints and personal
cloud drives. Safeguarding the security and
privacy of siloed information becomes a
complex, if not impossible, undertaking.

This only intensifies over time reaching a

nearly unmanageable state. Implementing
good data hygiene, consolidating where
feasible, and automating data management
tasks can significantly mitigate the issue
before it escalates into a crisis.
 CHALLENGE IV - REDUCE DISCREPANCIES BETWEEN TEAMS | 81

Stumble Street: Where enterprises

hit the speed bump
1. A clothing brand that needs to
standardize images and model poses
across its communications.

2. Unifying the disparate tech stacks at a

retail group with 37 mono sites and a
marketplace.
Smart fix:
Vue.ai also reduces discrepancies between
teams by:

• Automating document processing with IDP

to ensure consistency in data extraction and
verification. It also provides data integrity and
standardizes the processes.

• Automated tagging helps fix data. Clear,

standardized data processes streamline
operations, allowing teams to work in sync and
minimize errors.

• Vue.ai works across vendors and onboards,

enriches and organizes data over time, reducing
team complexity.

The homestretch:
A glimmer of optimism emerges with the advent
of connected journeys and federated intelligence,
facilitating seamless experiences that customers
crave for. Begin by acknowledging the multiple
personas interacting with an organization’s
touchpoints. Depending on the generation,
technological skill and other components that
define a demographic, approaching each persona
may require different strategies.

Also required: A sincere effort to dismantle the

discourse challenging the relevance of metadata
in the age of Generative AI. Even with AI actively
contributing to its assignment, metadata remains
pivotal to realize the true potential of AI. After all,
GenAI isn’t a one-size-fits-all remedy; it demands
discernment for crafting quality solutions.
 CHALLENGE IV - REDUCE DISCREPANCIES BETWEEN TEAMS | 83

The homestretch:
The question then isn’t as much of “What
is our AI strategy?” as it is about gaining a
profound understanding of the business
problem. Break free from the fixation on AI as
a universal panacea and redirect focus towards
understanding unique challenges. Then, cherry-
pick the right kind of AI tailored to specific
business issues, rather than blindly applying
software solutions.

The journey isn’t about clandestine cost savings

but a bold embrace of connected operations.
The time is ripe for enterprises to transcend
the silos, embrace the potency of connections
and orchestrate a symphony of customer
satisfaction and business triumph.
 ELEVATING
CUSTOMER
EXPERIENCE
 ELEVATING CUSTOMER EXPERIENCE | 85

Companies with the most mature digital

customer journeys see a 13.5% increase in
revenue and a 14% increase in profit, per an
April 2023 global survey of more than 1,500
leaders across nine countries and seven
industries. Successful enterprises are breaking
down the silos within their workforces and
data systems, providing customers with smart
recommendations and solutions so that less
time is spent on researching, selecting, and
ordering. Using pre-built, AI- and ML-based
demand planning technology, two-thirds of
the most mature enterprises have built a real-
time inventory ledger to provide a holistic view
across their networks. Businesses can now
analyze large data sets in the cloud and drive
improvements across the enterprise.

Clearly, a connected data platform that unlocks

real-time intelligence, enables workplace
collaboration and creates an immersive, AI-
driven personalized, customer experience is
becoming essential to long-term success.

Listening to our customers and understanding

their specific needs in this domain, we identify four
specific challenges faced by enterprises and have
designed solutions to effectively address them.
NOTES FROM
LEADERS

ADITYA
AGRAWAL

Head, Customer Experience Operations, TransUnion CIBIL India

 NOTES FROM LEADERS - ADITYA AGRAWAL | 87

With over 20 years of experience in the credit

information and lending industry, Aditya Agrawal
is a business management professional who leads
the customer experience operations at TransUnion
CIBIL, India’s leading credit bureau. He leverages
data, analytics, and technology to solve critical
business problems for credit lending and insurance
businesses, and to create a superior and enriching
experience for customers and partners.

As a digital transformation leader, he drives change

strategy and executes change management processes
to ensure customer success and sustainable business
profitability. He also leverages partnerships in the
business ecosystem to evolve rapidly and create new
business propositions. He has a keen understanding
of regulations and business environment, and has
built effective relationships with regulators, banking
ombudsman, and government authorities. Aditya
is an ambassador of the business, representing
the company across forums and public platforms,
and driving thought leadership through industry
representations and effective stakeholder
management. He is also an award-winning leader
who fosters a collaborative and high-performance
culture among cross-functional teams for business
growth and common goals.
CUSTOMER EXPERIENCE
FROM MY LENS
Aditya Agrawal,
Head, Customer Experience Operations, TransUnion CIBIL India

I have always believed that “If the why is stronger, the how
becomes easy” and this reflects in my everyday decisions.
The most important driving force for me and my team is a
strong belief that ‘Customer perception’ is our reality, and
this inspires us to keep the Customer at the center of all
experiences we design. In my business, artificial intelligence
is no longer a technology of the future, it’s today’s technology
shaping our tomorrow. It has become pivotal for us in
enhancing customer experiences, making them more
personalized, efficient, and emotionally attuned.

Experience which is meant for Everybody serves Nobody

As CX Experts, it’s our responsibility to harness AI in a way

that truly values and resonates with our customers as well as
the brand promise. Advanced analytics algorithms are helping
us sift through huge amounts of customer data, helping us
find patterns and insights humans may miss. This is helping
us identify customers’ wants, anticipate future needs, and
highlight areas where it matters the most for our customers.

From Multi-channel to a Seamless Omnichannel Experience

In my journey to execute the Customer Experience vision,

we are aiming for AI to unify data from different channels,
such as online, call center, and social media etc, ensuring
a seamless experience for our customers, who can switch
between channels and continue their journey without an
interruption.
 NOTES FROM LEADERS - ADITYA AGRAWAL | 89

Mix of Artificial Intelligence and Intelligent Automation

leading to designing Superior Experiences

The capacity of AI to analyze vast amounts of data,

understand customer behavior & preferences and
predict future trends has become an invaluable asset for
my businesses, considering the volumes we drive.

However, as we embrace these transformative changes,

our business tries to ensure that the technology serves
as an enabler to enhance human interaction, not replace
it. The aim is to automate the processes and decisions
that are repeatable and reproducible, helping us reduce
our error rates & response times and being first time
right. This leaves more time for my team to be creative,
and make decisions where human emotions and
empathy are required. This helps us balance the right
level of personalization and automation that we strive
to achieve, helping us deliver an engaging and tailored
experience for customers.
OUR FINDINGS
 CHALLENGE I - ENABLE CRO AND SMOOTHER WEBSITE JOURNEYS | 91

CHALLENGE - I

ENABLE
CRO AND
SMOOTHER
WEBSITE
JOURNEYS
Every detail of the passenger journey
– booking, when-to-leave, how-to-
get-to-airport, check-in, in-lounge and
in-flight experience – is optimized at
Qantas. Hundreds of partners offer its
customers novel ways to spend loyalty
points, according to their preferences.
Combining customer demographic and
past transaction data with social media
monitoring can help generate individualized
product recommendations. Next-product-
to-buy recommendations that target
individual customers – as companies such
as Amazon and Netflix have successfully
been doing – can lead to a two-fold
increase in the rate of sales conversions.

Interactions that our sales associates had

with enterprises followed a similar trail of
unique challenges and custom solutions to
meet them.
 CHALLENGE I - ENABLE CRO AND SMOOTHER WEBSITE JOURNEYS | 93

Stumble Street: Where enterprises hit

the speed bump
1. What good is a country with abundant
natural resources if it lacks the
infrastructure to make use of them? Or a
well-stocked library if no one understands
the language its books are in? A mass
market brand moving roughly 500,000 to
one million SKU’s a month was in a similar
quandary. They essentially wanted to
improve product discovery by bringing in
a layer of data enhancement that did not
exist before: By auto-tagging their images,
writing titles and descriptions and checking
if the product image matches the title.

2. In fact, a UK-based general marketplace

took this challenge right up to the doorsteps
of its vendors. By directly integrating
with the vendor’s portal, this digital-first
company mandated the approval of the
instantly-generated product tags and
descriptions by the vendors themselves,
facilitating a fantastic search experience for
the end users.

3. A franchise retailer with 1500 stores in

11 countries and another retailer with
about 60,000 SKUs per month – both
based out of the Middle East – faced
very similar challenges: Poor customer
experience arising from a disjointed array
of websites. Standardization of taxonomies
via unification of the tech stacks was the
critical need.

4. The capability to visualize products on

models is a very common customer
hankering among sellers in sectors as varied
as jewelry, apparel and accessories.
Smart fix:
Vue.ai’s Personalization Dynamo helps with
smoother website journeys by...

• Automating document processing with IDP

to ensure consistency in data extraction and
verification. It also provides data integrity
and standardizes the processes.

• Automated tagging helps fix data. Clear,

standardized data processes streamline
operations, allowing teams to work in sync
and minimize errors.

• Vue.ai works across vendors and onboards,

enriches and organizes data over time,
reducing team complexity.
CHALLENGE I - ENABLE CRO AND SMOOTHER WEBSITE JOURNEYS | 95

The homestretch:
A 15% boost to online AOV; 8% bump in
online revenues; complete automation of
product recommendations; 25% of sales from
personalized content. Furnishing and home
decor company Industry West achieved all
this by customizing its onsite content —
banners, hero images, special offers, and
product recommendations — to individual
audience segments. Irrespective of where
shoppers are in their purchasing journeys,
personalized content around what is most
relevant to them is what they see. Visitors
returning to the website pick up where they
left off while ‘new’ visitors receive a different
set of recommendations. Moving through
the buying process, product-specific content
that guides users toward an informed
purchase is surfaced.
 CHALLENGE II - CREATE AN OUTSTANDING OMNICHANNEL EXPERIENCE | 97

CHALLENGE - II

CREATE AN
OUTSTANDING
OMNICHANNEL
EXPERIENCE
Modern customers access contact centers
via nine or more channels. Weaving together
the structured and unstructured sources
from these multi-point sources is critical
to understanding customer needs and
crafting engaging experiences that are
mutually beneficial. Omnichannel data
leverages a customer’s unique preferences
and expectations for service, not just with
one brand but for every product and every
experience.

Think connected data is something only

large enterprises can handle? In 2021, the
33,000-stores-strong Starbucks that clocked
52% of its USA sales via digital channels was
upstaged by the 140-store restaurant chain
Sweetgreen (68% sales from online channels).
This it did with a cross-channel experience
built with data and digital in mind.
CHALLENGE II - CREATE AN OUTSTANDING OMNICHANNEL EXPERIENCE | 99

Stumble Street: Where enterprises hit

the speed bump
Here are some first-hand excerpts from our
sales leaders’ exchanges with practitioners in
the industry.

1. A large grocery retailer wanting to segment

its customers, to send targeted marketing
communications according to user
preferences.

2. A luxury brand manufacturer-cum-retailer

with around 5000 SKU’s looking to deliver a
personalized online experience.

3. Delivering a hyper-personalized buying

experience for insurance shoppers using
recommendations and lead scoring.

4. A fashion brand, a jewelry brand and a

retailer – all aspiring to improve product
search capabilities across channels and
personalize offerings to customers.

Smart fix:
Vue.ai’s Personalization Solution enables better
omnichannel customer experience through its
“Journey Orchestrator” & “Audience Builder”
tools. They help

• Craft and stitch personalized messages

across audiences, touchpoints, and
customer stages.

• Provide real-time customer behavior and

preferences data to optimize personalized
experience from start to finish.

• Businesses segment audiences into granular

cohorts that can be customized, enabling
easier targeting.
The homestretch:
Digital-advantage supremacy has now gone well
beyond the boundaries of traditional marketing
to become a much broader C-Suite issue. We are
now at the point where competitive advantage
will derive from the ability to capture, analyze
and utilize personalized customer data across
channels at scale and from the use of AI to
understand, shape, customize and optimize
the customer journey. Since personalizing
an end-to-end multichannel experience
requires orchestration across all the channels,
enterprises are collating multiple AI, martech,
and back-office solutions connected through
common-application programming interfaces to
better develop and use personalization data.
 CHALLENGE III - CRAFT A SEAMLESS ONBOARDING EXPERIENCE | 101

CHALLENGE - III

CRAFT A
SEAMLESS
ONBOARDING
EXPERIENCE
A 40% reduction in onboarding time for
automotive dealers and rideshare businesses –
without compromising on security and efficiency.
User identity verifications in mere seconds at
online gambling platforms. Banking conglomerates
reducing the time to open new accounts from 20
days to a matter of minutes – with 90% accuracy
and without compromising on compliance
obligations. AI powers smooth onboarding
experiences: From eliminating manual tasks for
buy- & sell-side teams in setting up custodial
and broker funds to verifying identities of tenant
applicants at rental property dealers.

We uncovered similar business needs at

companies across many industries.
 CHALLENGE III - CRAFT A SEAMLESS ONBOARDING EXPERIENCE | 103

Stumble Street: Where enterprises hit

the speed bump
1. Swift onboarding of BFSI customers through
verification of KYC documents and onboarding
staffers in the healthcare industry according to
specified prerequisites.

2. Automating the onboarding of restaurants

by food delivery services: Auto-verification
of onboarding documents such as licenses,
contracts, menu cards, cuisine tags, bank
details, CR, MIF, VAT.

3. Authenticating the validity and identity of new

/ repeat visitors at the gate pass systems of a
business enclave, based on their Identity Proof
document.

4. Scoring leads and opportunities at a US-

based medicare provider; identifying the right
products for each lead and mapping them to
the right sales agent; prioritizing a lead with a
higher chance of converting vs an opportunity
with a lower chance.

Smart fix:
Vue.ai’s Intelligent Document Processing solution
significantly enhances the onboarding process by:

• Ingesting and extracting information from various

document sources, including PDFs, images, and scanned
documents. This automation streamlines data entry,
reduces errors, and accelerates the onboarding timeline.

• Improved speed, accuracy, and overall efficiency in

handling diverse document types during the onboarding
process ensures the accuracy and consistency of
extracted data, enabling efficient cross-verification for
compliance purposes.
The homestretch:
Rolls-Royce can take a year or longer to qualify,
test, certify and onboard a new supplier –
but that’s all in the past. Thanks to AI, today
fintechs, rental and hiring companies, gaming
and government organizations provide
friction-free, compliant and quick customer
verification processes. Later this year, when
the value of AI-underwritten insurance premia
will top USD 20B (Juniper Research), faster
customer onboarding will no wonder be a top
contributing factor.
 CHALLENGE IV - ACTIVATE A FANTASTIC CUSTOMER SUPPORT CHANNEL | 105

CHALLENGE - IV

ACTIVATE
A FANTASTIC
CUSTOMER
SUPPORT
CHANNEL
Across Marketing, Sales and support
functions, AI-facilitated chatbots have
saved business leaders some $300,000
a year. And that was even before the
pandemic! To understand customers better
and faster than humans alone could, use AI.
Insights collected, analyzed, and processed
by AI in real-time can augment the work of
Customer Support services, helping them
rationalize where and how to dedicate time
and effort. Generative AI can also automate
some of their simpler functions, to open
more time to drive meaningful engagement
with brand communities. It’s how the
television series ‘Atlanta’ let a custom AI
bot take over the show’s official Twitter
account for a week last year.
CHALLENGE IV - ACTIVATE A FANTASTIC CUSTOMER SUPPORT CHANNEL | 107

Stumble Street: Where enterprises hit

the speed bump
An insurance firm that required a Generative
AI tool to query into their knowledgebase of
documents (PPTs, PDFs, Images, etc.) to find
answers on the fly to customer’s questions.

Smart fix:
A robust Customer Support channel can be
elegantly enabled by leveraging AI in the following
key aspects.

• Streamlined product onboarding: AI can

enhance customer support by facilitating the
onboarding process, meticulously guiding
through the completeness and accuracy
of data sets, spontaneously flagging any
inconsistencies, duplication, or deviations
from the business’s established standards,
and empowering continuous feedback
mechanisms.

• Data-driven personalization: The activation

of AI allows for the provision of real-
time analytics on customer behavior and
preferences. This invaluable data is utilized to
fine-tune the support experience, customizing
interactions and solutions to meet customer
needs from the outset, thereby optimizing the
overall customer journey.

• Integration of recommendations: AI empowers

customer support channels by placing tailored
recommendations throughout the website
judiciously, making the browsing experience
more intuitive for the customer. Additionally,
it plays a strategic role in molding business
workflows to be more efficient and responsive
based on insights garnered from customer
purchase patterns.
The homestretch:
Integration of AI, automation and advanced
analytics in customer service is proving to be
a significant driver to providing consistent,
cost-effective and personalized support
to customers – especially with seasonal
fluctuations and high expectations during peak
shopping seasons. Signature Kitchen Suite uses
AI to analyze appliance usage patterns, provide
contextual reports on product performance,
contact the homeowner should a potential
service issue be detected or if supplies are
running low and give guidance on how to best
use its product range. Customer support for
mobile-first tech industries like entertainment,
fintech and gaming already see automation
rates of 82-94% for resolved issues.
 109

Innovation distinguishes
between a leader and a follower.
— Steve Jobs
ABOUT VUE.AI

The only AI stack you will ever need

Vue.ai is an enterprise platform that engineers bespoke AI
transformation roadmaps across industries and delivers business
outcomes – obsessively. Retailers to resellers, auto-extracting data
from files to extrapolating fashion styles, 150+ conglomerates in five
continents recruit Vue.ai.

Vue.ai’s Orchestration Platform

Plug all your apps and data into a single source, and watch AI
power your field, IT and analytics teams.
 ABOUT VUE.AI | 111

An AI walks into your board meeting

From the board to your peers, everyone wants quick
outcomes. Our 30:60:90 promise puts you on the course for
your first, quick AI ROI win, within 90 days.

QBR? Ready, we are.

Global enterprises powered by Vue.ai

ABOUT VUE.AI
 RESOURCES | 113

RESOURCES

Events

REBUILD: The AI
Transformation Retreat

Blogs
FOUNDERS’ NOTE:
From Noise to Signal: A 2024 AI Memo

Podcasts
Go grocery shopping with
a dash of data and AI.
Podcast with Kirk Ball, EVP
& CIO, Giant Eagle, Inc
Write us at [email protected]