Unlocking Growth - Introduction to GenAI - Part 1 of 2

Unlocking growth - by leveraging the power of AI and GenAI

1. WwW N i

Growth

enerative Al

tion to if

2. Contributors...

Murat! SUNDARAM L SHRINIVAASAN SRIRAM H CHIRANTHAN RAM

MANIKUMAR SV SIVAKUMAR VAIDYANATHAN C VIJAYAKKUMAR S

HARIBASKARAN R KOUSHIK

© 2028. NetworkGain Consulting.

3. Information Source...

365 DATA SCIENCE ADOBE ATRIBS AMAZON

BAIN & Co BOSTON CONSULTING BUSINESS INSIDER Cap CORP

CLAUDE DATA SCIENCE Doso DELOITTE FORBES

FREEPIK GARTNER GOOGLE HAVARD BUSINESS REVIEW
IBM MICROSOFT NETWORKGAIN CONSULTING OPENAI

XMPLAR

© 2024. NetworkGain Consulting.

4. Imagine...

Food for thought

© 2028. NetworkGain Consulting.

5. The Time Saver

Q: Imagine you've a task that takes 10
hours to complete manually. You find a
tool that automates 80% of the task. How
much time will you save, and how can you
use the saved time?

TIME given

to THOUGHT is
the GREATEST
TIME SAVER
of ALL

Y A: You will save 8 hours (80% of 10 hours). With those 8 hours, you can focus on other
.un.. OOOO OOO important task or spend time doing things you enjoy.

6. The Data Deluge

Q: You've a dataset with 1 million rows of data.
Analyzing this data manually would take
weeks. If an Al tool can analyze and generate
insights in just few minutes, how much faster 9 OVERLOADING
is the Al, and what does this imply? ? Pr

PR
ou”

A: If manual analysis takes say 3 weeks or 504 hours, Al can do it 10 minutes, that is u —
3,024 times faster (504 * 60) / 10.

7. The Repetitive Tasks

Q: You perform a repetitive task every day
that takes 30 minutes. Over a year, how
much time does this task take, and how
could automating it with Al change your
daily routine?

A: (30 minutes x 365 days)/60 = 182.5 hours. It is almost a month of free paid time.

8. Facts and Stats

With a dose of Al humor

© 2028. NetworkGain Consulting.

9. *

Al takes world by storm

By 2025, the global Al market is

expected to be worth over S 190 bition!
That's almost enough to buy everyone on

Earth a cup of coffee... if Al doesn't drink it
first.

10. alone, the market for Gen Al

tools grew by over Looks like Al
isn't just generating content, it's generating
profits faster than you can say “deep
learning”

* %*

Gen Al's explosive growth

11. ? *

Al's creative streak

OED

lin 5 companies now use Al to generate

content, from blog posts to music. Who
knew the next Picasso might be a server ina
data center?

© HE |

12. *

Al in the Workplace

72% of executive believe Al will offer significant

business advantages. The other 28%? They're still
trying to convince their computers that Excel is not
Al

13. Everyday, Al model process roughly 2.5
quintillion bytes of data. That's like reading
the entire U.S Library of Congress 15,000
times a day. Clearly, Al has no time for cat D a
videos (unless it's learning to identify them). !

JIE

S
OF
o

Al is always learning

14. Al loves efficiency
Al-driven automation could increase global

GDP by up to 915.7 trillion by 2030.
That's like adding another economy the size
of China, but with fewer lunch breaks.

*

15. Gen Al vs. Creativity

Over 60% of creatives now use Al to

assist with their work. Al might soon write
the next bestseller, but it still hasn't figured
out how to take a break

{0} ==
S Ss

16. Al ethics - a growing concern

63% of people worry about Al's
impact on privacy. The other 37%? -

they're probably Al bots themselves

9
A=

?
om

17. *

| u Al in Healthcare 2

>) Al could reduce healthcare costs by
a S 150b annually by 2025. Just what the

doctor ordered minus the waiting room
magazines

*

18. XX

Al's social skills

In 2024, its estimated that 90% of
customer interactions will be managed by
Al. So, if your next customer service rep
sounds suspiciously friendly and never gets
tired, you know why!

19. Speak

the tech language - you'll find the code to unlock innovation

© 2028. NetworkGain Consulting.

20. What is?

Let's understand the building blo

21. Artificial Intelligence (Al)

giving computers a
brain to make
smart decisions

25. n(B) =68
n(c) =84

NI e

y) = log.x + logny
al = log,x- logy
A B albrc)=abrec 100002+100b-

126 =6xy a= +.
2x+2y=20 2"

¡AAA

26. rstand, learn
erform

29. Unleash

the Power: Generative Al at Work - Transforming Ideas into Innovation

© 2028. NetworkGain Consulting.

30. Leading business executives - uses
GenAl

‘Sundar Pichai, CEO of
Google, said he used a
language model to talk to
the planet Pluto with his
son.

Venture capitalist Vinod

Khosla said he used
ChatGPT to write a rap for

AMD CEO Lisa Su said she a eso

uses Microsoft's Copilot to

summarize meetings and

track actions. CEO and cofounder of
OpenAI Sam Altman said
he uses his company’s
chatbot for translation and
writing.

© 2028. NetworkGain Consulting.

31. Evolution of Al

Evolution of Al

Artificial
Intelligence

Deep
Learning

Machine

—
Learning

Generative Al
——

Evolution of technologies like cloud
computing and Big Data have enabled
better Al capabilities

Advances in Computer Hardware such as
GPUs and larger storage have also
contributed to rise in Al

While Machine Learning & Deep
Learning techniques can predict future
outcomes, Gen Al can create new
content based on this prediction

32. GenAl Consumer Web Products:

4 @chacer N. IlElevenLabs 21. PhotoRoom 31. PIXA! 41. +! MaxAlme
2 Gemini’ 12. [El Huggingrace 2. YODRYO 32. ideogram 42. f Craiyon
3. character.ai 8, @ Leonardo.ai 23. a Clipchamp 33. @invideoat 43. (D opusciip
4 «liner 14. PN Midjourney 24. (8 runway 34. [Feplicate

s. MauillBot 15. © SpicyChat 25. YOU 35.) Playground 45. @cHATPDF
6 Poe 16. @Gamma 26. DeepAl 36. #Suno 46. — MPIXELCUT

7. Gf) perplexty m. Qorusona 2. (JEightify 37 Y) chubai a7. E Vectorizer Al

8. JanitorAl 18. cutoutpro 28. candy.ai 38. Speechify 48. CDDREAM

9. CIMITAI 19. QPIXLR 29. Might Cafe 39. phind 49. Photomyne

10. Claude 20. VEED.IO 30. VocalRemover 40. 4 NovelAI so. Ollfes Otter.ai

33. GenAl Consumer Web Products:

CIITA menu
Meplicate

Bow Claude
DeepAl Gemini
‘ Poe

‘candy.ci charocecal

Dom Oman

De anita
Smee JOAO.

34. GenAl create outputs

(Y)
NA

Code

35. GenAl Application Landscape

36. GenAl Market Map + Infrastructure Stack

37. Key benefits of GenAl / LLM

Powerful tools for transformation

nAl and LLM are pow
aspects of business ai

ful tools tl

driving innovation, efficiency,

at can transform various

1 2

Automated content Automation of repetitive
creation tasks

3 5 6

0 Predictive Analytics Customized user New Product and Service Adaptable and Scalable
experiences Development Solutions

and personalized experiences

7 8 9 10

Continuous Improvement | Fraud Detection and Enhanced Accessibility Sustainability Initiatives
Prevention

38. GenAl

Reshaping the Future

Other Al models and
frameworks, including near-AGI

© 2028. NetworkGain Consulting.

Types of Al enablement

Generative Info
collection
—t
Generative
research

Generative
insights
Generative

Generative
innovation

a
Generative
decisions

Al-Enabled
knowledge worker

Al
7

Strategic
Work

Al-Enabled
Work Apps

Tactical
Work

How Artificial Intelligence will
reshape the workplace and
employee experience

Pursuing new markets
Designing CX/EX experiences
Imagining new products + processes
Creating + using unique knowledge
Communicating + collaborating
Making better decisions, faster
‘Optimizing business operations
Gathering needed information

39. How AI helps business process?

What should Customers

be my market

x
poston
strategy? à

s |

‘Operations staff

Planning
Strategy

my available

capacity? What
are open

payables?

Operations
Management

suppliers

ERP

=n

a
—
LY
wer

rares Scesuirg &

Ditton route optimization

Sales HR Finance Marketing
oscensten || stenentrageren | | -aoimocecasnce | | -omentemmung
rocas | | !erpaeedene a | ces

a I See Ser get
Germercatons | | "Berner rento] [segments
E
Manufacturing Al provides an interface between
svete on entrer

businesses and traditional operational
systems

40. How people are
using GenAl

+++++++++++++
‘Thomes DE 22222227
2222522255

1
228222222222
+++++++++++
+++++++++++

© 2024. NetworkGain Consulting.

© Viga complaint

CEE

© troncs

Coding for amateurs

© Wenig senmnes
Cerna

© botines

© Stig

© Crain tty any

© (ngs ep dosent

O Business advice

© retorna
Canin ct po

ti ps mar et

© canin sen pan

© Messie des!
© Focecheching
© Cover advice

© wows oer oes
"rca move pts

Crab we pe

© reine rascal mgens

© trote.
natal menea

O boton dons

O som
Sums code tans

© Pacing tet

Rater ching sugngcosa Sampling data

PAS
(O For people with ADHO
© éme om

© Preso comenten) Seca et ean

© Seveg died spots
© vn

O topo reses,
© Wing posts

OD mcrnnendmone bois ac E) Srengreringan met
Cooking with what you have D) Jumping tothe vst info

© Biting bis lan

eto prompt

For eternas

© But a vebsteps
Wine og post
Wing eng proposal

© Wing press los

© Fag ra ages

© Project management

40

41. Indian entities adopting GenAl

Company Name
Tata Motors
BigBasket

| Ola Cabs
Lenskart
Swiggy
Zomato
Udaan
PolicyBazaar
UrbanClap
Pepperfry

Use Case
Predictive Maintenance
Personalized Recommendations
Dynamic Pricing
Virtual Try-on
Delivery Route Optimization
Customer Support Chatbots
Inventory Management
Fraud Detection
Personalized Service Recommendation

Visual search and recommendations

Al Adopted
IBM Watson
Google Cloud Al
Microsoft Azure Al > Krutrim Al
DeepAR Al
In-House Al
Haptik Al
AWS Al
TensorFlow
IBM Watson
ViSenze Al

42. Vertical-wise Use Cases

HEALTHCARE
- Personalized Medicine
Healthcare Chatbots and Virtual
Assistants
- Clinical Data Generation
- Drug discovery & personalized medicine
- Patient-specific treatment plans
- Medical imaging analysis & diagnosis

RETAIL AND E-COMMERCE
- Product Image Generation

Virtual Try-On

Dynamic Pricing Optimization

Content Generation

Personalized product recommendations
8 marketing campaigns

Demand forecasting & inventory

- Virtual clinical trials & drug development management

Personalized health education & support

ENTERTAINMENT AND MEDIA
- Content Creation
Deepfake Detection
- Music Generation and Composition
Story Generation
- Virtual Production

Dynamic pricing & product optimization
- Product description & review generation

Chatbots for customer service & product
recommendations

TRANSPORTATION

- Personalized route recommendations
Self-driving cars

- Predicting maintenance needs
Personalized travel recommendations

- Al-powered logistics systems

FINANCE AND BANKING
- Fraud Detection and Prevention
Algorithmic Trading Strategies
Credit Scoring and Lending
Financial Forecasting and Risk
Management
- Anti-Money Laundering (AML)
Compliance
- Algorithmic trading & portfolio
optimization
Financial report & analysis generation
Personalized financial advice & planning
Chatbots for customer service & loan
applications

AGRICULTURE

‘Optimizing crop yields
Detecting and predicting plant diseases
- Automated harvesting robots
Personalized fertilizer and pesticide
recommendations
Predicting market trends

MANUFACTURING AND ENGINEERING
- Generative Design

Virtual Prototyping and Simulation
- Supply Chain Optimization
- Quality Control
- Production process design &
optimization
- Predictive maintenance & quality control
- Personalized product manuals &
instructions
- Logistics planning

Developing new materials & product
prototypes

ENERGY & ENVIRONMENT

- Designing and optimizing renewable
energy systems
- Predicting energy demand

Monitoring environmental changes
- Al-powered disaster management tools
- Personalized energy consumption
recommendations

43. Functional Use Cases

ISTOMER SERVICE AND DATA ANALYTICS AND
SALES HUMAN RESOURCES MARKETING AND ADVERTISING papain ren
- Lead Generation - Resume Generation - Creative Content Generation - Chatbots and Virtual Assistants _- Data Augmentation and
- Content Creation - Candidate Screening - Personalized Marketing Natural Language Generation Synthesis
Sales Forecasting Interview Simulation Materials Sentiment Analysis Predictive Analytics
- Customer Support Employee Training AVB Testing Optimization Call Center Automation ‘Anomaly Detection
- Personalized Recommendations - Diversity and Inclusion - Social Media Content Voice Synthesis Customer Segmentation and
- Sales Training - Employee Engagement Generation Targeting
Competitive Analysis Performance Management Influencer Identification and - Market Research and Trend
- Automated Email Campaigns. - Succession Planning Recommendation Analysis
- Dynamic Pricing - Employee Well-being

- Sales Performance Analysis

RESEARCH AND DEVELOPMENT CONTENT CREATION DESIGN 8: ENGINEERING DATA ANALYSIS Se SCIENCE
Drug Discovery and Design Personalized marketing copy & ads Product design & prototyping Synthetic data generation

- Protein Folding Prediction Creative writing & storytelling Code generation & documentation Pattern & trend identification

- Material Science Optimization - News & report generation - Engineering process optimization Predictive modeling & forecasting

- Chemical Reaction Prediction - Music composition & audio generation - Predictive maintenance & anomaly - Drug discovery & material science

- Automated Literature Review Image & video generation detection simulations

- Material science simulations & discovery - Personalized education & learning
materials

44. Understanding TCO

+ Ingredients (Building Blocks)
+ Lemonade Recipe (Al !
+ Sugar, Lemons, and Water (Baits)
* Cups and Straws (Computing Power)
+ Helpers (Services)
+ Lemonade Makers (Software
+ Servers (Support and Maintenance
« Location (Infrastructure)
* The Stand (Cloud Services)
+ Lemonade Price (Pricing)
+ Per Cup (Per Use)
+ All-You-Can-Drink (Subscription)
+ Keep Everything Running (TCO)
+ Buying mode lemons (Ongoing costs)
+ Fixing the Stand (Maintenance)

45. TCO = One Time + Fixed + Variable Costs

=(YHo2>+: dol > + Pog

TCO = (InmaL Al MODEL + INFRASTRUCTURE SETUP) + {REGULAR CLOUD SERVICE FEES + SUPPORT SALARIES} + (USAGE-BASED COMPUTE POWER + SCALABILITY}

ONE-TIME COSTS (INITIAL FIXED COSTS (REGULAR VARIABLE COSTS (DEPENDS ON

INVESTMENT) PAYMENTS) USAGE)

+ BUYING THE RECIPE + MONTHLY RENT * INGREDIENTS PER CUP

+ SETTING UP THE STAND + SALARIES FOR HELPERS + EXTRA HELPERS DURING RUSH
HOUR

45
© 2028. NetworkGain Consulting.

46. Start

Your journey to Innovation begins now...

© 2028. NetworkGain Consulting.

47. Ready, Get Set, Go!!!

48. Challenges / Risk in adopting

> as Ethical

dilemmas

Malicious

use of

generated | Cyber
data security
and
privacy

Intellectual un & Fake content propagation
property & =

49. Responsible Al framework

Data & Al Ethics
Consider the moral
implication of uses of
data and Al and codify
them into your
‘organization's values.

Policy & Regulation
Anticipate and
understand key public
policy and regulatory
trends to align
‘compliance processes.

Governance
Enable oversight of

systems across the three.
lines of defense.

Compliance
‘Comply with regulation,
‘organizational policies,
and industry standards.

Risk Management
Expand transitional risk
detection and mitigation
practices to address
risks and harms unique
to AL.

Interpretability &
Explainability
Enable transparent
model decision-making.

Sustainability
Minimize negative
environmental impact
and empower people

Robustness

Enable high performing
and reliable systems.

Bias & Fairness
Define and measure
fairness and test systems
against standards.

Security
Enhance the
cybersecurity of systems.

Privacy
Develop systems that
preserve data privacy.

Safety
Design and test systems.

to prevent physical harm.

Problem
Formulation

Identify the concrete
problem you are
solving for and
whether it warrants
an AUML solution,

Standards
Follow industry
standards and
best practices.

Validation
Evaluate model
performance and
continue to iterate on
design and development
to improve metrics.

Monitoring
Implement continuous
monitoring to identity
dit and risks.

51. Use Cases

52. To integrate Al into
an existing Identify Business Set Clear Assess Data
business, consider Processes: Objectives: Availability:

the following steps:

—

53. Prompt Mastery

WHAT

* What are you asking the system to
do?
* What is your desired output?

“Please show me the steps to implement a

Random Forest Algorithm in Python. My desired

output is a comprehensive step-by-step guide to
carry out this task’

WHERE

* Where in terms of industry, sector,
geography?

“The goal is to build a robust ML model. This
prompt is critical as the implementation of a
successful Random Forest algorithm can
enhance our products effectiveness and
precision in risk prediction”

© 2028. NetworkGain cı

WHO

= Who is asking?
= Who should respond?

WHY

* Why do you ask the question?
* Why do you need the output?
* Why is this an important prompt?

“The question originates from a fintech startup
located in Chennai, Tamil Nadu. This startup
specializes in building Al-driven tools for risk
assessment in the financial domain primarily

serving banking institutions across India”

"Im a beginner data scientist. Ideally, please
answer as on experienced data scientist.”

Mos
= Which are the relevant details? = How should the Al respond?

= How to structure the desired output?
= How extensive / concise?

“we need a detailed response. Please include a
step-by-step guide to building Random Forest
model in Python, with code snippets and
explanations. it should also include guidance on
parameter tuning and interpreting the results.
The output should be structured in a way that it
facilitates easy understanding and seamless
implementation

“we've a mix of numerical data (such as account

balances, transaction volumes, loan repayment

rates) and categorical data (like loan type, credit
ratings)”

54. Tips for Prompting

ANALYZE the request for proposal
(RFP) requirements.

ANALYZE the RFP for the
customers top priorities,

points, size, scope, complenity, and

‘ANALYZE the evaluation criteria,

term and conditions, language and

ARGUE persuasively for the
superiority of our solution

ASSESS the potential risks
associated with the RFP

risk to operations. rome TEMES
BRAINSTORM innovate CATEGORIZE the key features of CLARIFY the scope and objecives COMPARE our staffing strategy to CONCLUDE by highlighting the
approaches to the solution the product offerings Of this proposal” ‘competitors in the market eue)
CONTRAST the benefits of the CONVINCE the client of the ROI CREATE a compelling vision for CRITIQUE potential drawbacks in DEFINE the project scope,
solutions with those on the market they gain the projects success ‘the approach deliverables, and timelines
DEMONSTRATE how our solution DESCRIBE the key features and — DETAIL the implementation plan DEVELOP a customized solution DISCUSS the market trends
solves the clients’ challenges benefits of the offering for project execution tailored to the client driving the need for the solution
DRAFT a comprehensive proposal ELABORATE on the innovative ESTIMATE the cost and time EVALUATE the solutions’ potential EXAMINE the current challenges
for client review aspects of the approach required for project completion impact on the clients’ business. face by the client
EXPLAIN the technology stack and FORMULATE a clear strategy for
its advantages Project implementation

© 2028. NetworkGsin Consul

55. Get Your Hands on the Future: