ai-powered-tech-services-a-roadmap-for-future-ready-firms.pdf

1
AI Powered Tech Services: A Roadmap for Future Ready Firms
February 2024
AI POWERED TECH SERVICES:
A ROADMAP FOR
FUTURE READY FIRMS
AI's Role in Turbocharging the Industry

2

Boston Consulting Group partners with leaders in business and
society to tackle their most important challenges and capture
their greatest opportunities. BCG was the pioneer in business
strategy when it was founded in 1963. Today, we work closely
with clients to embrace a transformational approach aimed at
benefiting all stakeholders—empowering organizations to grow,
build sustainable competitive advantage, and drive positive
societal impact.
Our diverse, global teams bring deep industry and functional
expertise and a range of perspectives that question the
status quo and spark change. BCG delivers solutions through
leading-edge management consulting, technology and design,
and corporate and digital ventures. We work in a uniquely
collaborative model across the firm and throughout all levels of
the client organization, fueled by the goal of helping our clients
thrive and enabling them to make the world a better place.
NASSCOM represents the voice of the $250 Bn plus
technology industry in India with the vision to establish the
nation as the world’s leading technology ecosystem. Boasting
a diverse and influential community of over 3000 member
companies our network spans the entire spectrum of the
industry from DeepTech and AI start-ups to multinationals
and from products to services, Global Capability Centres
to Engineering firms. Guided by our vision, our strategic
imperatives are to accelerate skilling at scale for future-
ready talent, strengthen the innovation quotient across
industry verticals, create new market opportunities - both
international and domestic, drive policy advocacy to advance
innovation and ease of doing business, and build the industry
narrative with a focus on Trust, and Innovation. And, in
everything we do, we will continue to champion the need for
diversity and equal opportunity.

4

5
TABLE OF CONTENTS
The AI & GenAI
Market Landscape
10-25
01
AI Acceleration
Framework & Imperatives
for Tech Services
26-61
02

6
EDITOR'S NOTE
We live in an era where Artificial Intelligence (AI) has not only invaded our
living rooms but has also ascended to a place of strategic prominence in
the global business arena. More importantly, its ascent has been swift as
evidenced by the fact that ChatGPT took only 5 days to reach 1 Mn users. In
contrast, just a few years back, Instagram and Spotify took 75 days and 150
days,respectivelytoachievethesamemilestone.TheAISoftware&Services
market is already valued at c. $100 Bn and is expected to reach $300-320 Bn
by 2027. Further, investments in AI are also booming across the globe with
a sizeable $83 Bn invested in 2023. Interestingly, data & analytics and
Gen AI emerge as dominant themes with the former attracting investments
worth c. $42 Bn and the latter c. $23 Bn.
Inarguably, the spotlight is now on technology service providers. Amidst
this backdrop, Indian tech giants and nimble mid-scaled players, along with
BPO and GCC stalwarts,are racing to harness the potential of GenAI,pouring
investments into crafting innovative solutions. But the question to ponder
upon is, “Are tech services players well equipped to meet the AI and GenAI
needs of their clients?”
In this collaboration, BCG and NASSCOM delve into this pressing question,
employing a comprehensive six-dimensional framework to evaluate the AI
maturity landscape across India's tech ecosystem which spotlights:
● AI for clients
● Vision & governance
● Operational model
● People
● Technology
● Data
The report aims to illuminate the path taken by exemplars to serve as
learnings for the sector – highlighting their key achievements in AI and
defining the strategic actions for the way forward.
Exemplars have developed a wide array of innovative AI-based services
and solutions, expanding their offerings beyond traditional IT services. This
RAJIV GUPTA
MANAGING DIRECTOR
AND SENIOR PARTNER
BCG
DEBJANI GHOSH
PRESIDENT
NASSCOM

7
includes the development of proprietary AI & GenAI platforms, tools for
automation,data & analytics solutions,and bespoke AI applications for
specific industry verticals such as healthcare, banking & finance, and retail.
30% of players have also expanded further to offer GenAI advisory and
custom model finetuning services for domain specific solutions. For
instance, an exemplar mid tech player is offering a GenAI platform as a
service for its clients which enables them to finetune & test models and
develop apps for different use cases. This has enabled the firm to deepen
market penetration and open new revenue streams.
25% of the companies have been able to build a sizeable 20%+ AI and 5%+
GenAI linked client portfolio. On GenAI specifically, they have been able
to scale GenAI use cases to production – exemplars have seen up to
10 use cases in production, vs industry average of 4-5. Some examples of
these productionized use cases include customer experience, marketing
content generation, GenAI-enabled SDLC, which allow for replicability
across sectors as well as sector-specific use cases like claims management
& fraud detection. E.g., a BPO player built a claims management platform
for an insurance firm by deploying AI+GenAI algos at every decision point
of the value chain (adjudication, payment calculation, litigation, and others).
Further, 45% companies have significantly enhanced their operational
efficiency with 20%+ realized efficiency seen in pilot accounts of
Application Development.
There are many success factors on how the exemplars have been able to
propel their AI journey:
● Firstly, leading firms have recognized the importance of human
capital in the AI journey and accordingly invested heavily in upskilling
and reskilling their workforce in AI and related technologies, with some
allocating $1 Bn over the next 3 years to upskilling.There has also been
a corresponding increase in the demand for AI skills - AI leadership
hiring saw a 15% increase while AI engineers’ hiring rose by c. 70%
in the last year. While ML, Python & SQL continue to dominate current
skills requirement in AI, GitHub, PyTorch and Databricks are emerging
as important skills as well. The right technical expertise acts as a
key differentiator for these players, enabling them to converse with
potential & existing clients.
● Secondly, players have established AI Centers of Excellence (CoE)
with a dedicated leader for driving AI business for the organization.
These CoEs are x-functional and x-sectoral and work seamlessly
with BUs, involving representation from technical, business, and legal.
They have:
• Dedicated AI SPOCs working in tandem with BU heads to build
commercially driven AI solutions and are responsible for ensuring
• Holistic viewpoints when building & prioritizing use cases by
evaluating business impact x implementation feasibility x
ability to replicate and scale x risks.

8
● Thirdly, exemplars today have moved ahead of the pack in terms
of building and adopting a unified GenAI platform with ready
integrations with common LLMs and data ingestion & labelling
methodologies. This allows them to rapidly build and deploy smaller
PoCs – shrinking from a timeline of few months to few weeks –
based on client needs.
● Fourthly, a common GTM strategy employed is paid soft launches that
provide comfort and early results to clients. These soft launches allow
players to ensure the seriousness of intent and foster a co-creation
process with clients. This has allowed companies to grow their AI &
GenAI client portfolio with approximately 15% firms expected to add 20+
new clients in the next year.
● In addition, tech services firms are forming strategic alliances with
not only the tech giants but also with niche startups for specific
horizontal (e.g., knowledge management – writer) themes.
This brings us to some key questions - what does this all mean for players in
the market and what can the tech services industry in India learn as it looks
to further its AI agenda? Here are 3 key imperatives for players to keep in
mind as they accelerate their AI journey:
● The biggest gap lies in understanding what clients are willing to
experiment with vs what they are willing to pay for vs what can generate
the highest value at scale. Therefore, it is critical to have an account
wise defined GTM strategy with differentiated and customized offerings
based on an in-depth understanding of client needs and areas where
our ability to serve them is the highest. Ability to create accelerated
PoCs for clients via a structured platform/framework is key for tech
services players.
● In addition to getting the right technical expertise via a structured
people strategy, business acumen applied to AI & GenAI use cases is
an important talent imperative. In order to bridge this gap, the focus
on upskilling in AI & business acumen should shift from not just
the delivery teams to sales & pre-sales as well as internal functions,
amongst others. Investing in and inculcating consultative-led selling for
AI offers can act as a key differentiator.
● A fundamental rethink of the Op model structure is imperative with
AI & GenAI use cases leading to new delivery structure design as well
as multi-departmental collaboration on GTM motions & solutioning.
The internal AI agenda needs to function as a living entity – constantly
evolving to fit the rapid changes in the tech landscape with robust
prioritization practices in place.

9

10
01
The AI & GenAI
Market Landscape

11
Key highlights of AI & GenAI Market Landscape
Demand for AI talent growing at c. 15% in India
● India has the highest AI skills penetration with 3x more AI skilled talent than other countries; Over the last 7 years, India has
witnessed a 14x growth in individuals skilled with AI
● While demand for talent in India is expected to grow ~15%, the market is expected to grow 25-35%, indicating the need to focus
on upskilling existing talent as well as breaking the linearity of growth in revenues and # of FTEs
● Tech services firms have begun to actively hire for AI/ML specific jobs: AI leadership hiring increased by c. 15% while the hiring
for AI engineers rose by c. 70% in the last year
● While ML, Python and SQL continue to dominate current skills requirement in AI; GitHub, PyTorch and Databricks are also
beginning to emerge as important skills
Perspectives
on Indian AI
Talent
The global AI market is expected to grow to $320-380 Bn by 2027 witnessing a CAGR of 25-35% with software & services segment
expected to account for c. 88% of the market.
In 2023, the 'Sandboxing Into Future: Decoding Technology's Biggest Bets' report identified AI/ML as one of the biggest
technological disruptors. This report delves into AI/ML’s potential to disrupt the tech services industry.
● GenAI is expected to comprise c. 33% of the global AI market by 2027 while ML & Vision will comprise another c. 29%.
● The financial sector will continue to be the highest contributing sector followed by media & entertainment
● In line with this growth trend, IT buyers expect to increase their spend on AI, cloud & analytics in the forthcoming year; Spend on
server infrastructure expected to reduce the most
● While India’s AI market is likely to grow on par with the global market with a skew towards the financial sector as the main
spender, the tech sector is also expected to increase spending in India over the next few years
AI Market
Overview
Investments in AI are booming across the globe with $83 Bn invested in 2023. Key themes emerging:
● Data & analytics emerges as the dominant theme ($42 Bn)1
, with GenAI ranking second ($23 Bn)1
. This suggests the expectation
in value creation from serving enterprises in setting their data architecture and training data in place, so they can leverage the
full benefits of GenAI
● Tech ($58 Bn)1
, banking ($27 Bn)1
& healthcare ($23 Bn)1
emerge as the top 3 sectors to receive AI funding globally
● While data & analytics remains the highest across regions, GenAI investments are largely skewed towards NAMR currently, with
c. $30 Bn invested in 2 deals (Nuance, OpenAI). India is seeing players emerge, e.g., Sarvam.ai raised $40 Mn in Series A
● Disproportionate investments seen in HR/CRM in Europe and in marketing & advertising solutions in India
● c. 55% of investments made by Indian tech products & services investors have been in India itself in the form of strategic bets;
Top 3 deals account for c. 58% (Jio investing in Glance, Perfios in Karza, and Infosys in SAFE Life Science) of the investments
AI
Investments
Landscape
Note: 1. Investments can span across multiple horizontal themes and verticals

12
Recap | Our last report identified AI/ML as one of the key Tech disruptors;
This report deep dives on AI/ML's potential in disrupting Tech Services
1. Technologies in bottom 30 percentile of funding kept in innovation 2. Freshness-% of patents filed in last 3 years
Note: Funding data based on publicly disclosed deals and reflects private investments for applications of a technology
Source: BCG Analysis, Sandboxing into the Future: Decoding Technology's Biggest Bets - NASSCOM-BCG Report, Dec'22
Innovate Incubate Commercialized
Funding
Momentum
(50%-funding
in
last
5
years,
50%-CAGR)
Wireless
Low Power
Networks
Cognitive
Computing
Gesture Recognition
Outdoor Location Intelligence
Serverless
Computing
Web 3.0
Quantum Computing
Virtual Agents
Industry Cloud
Metaverse
Edge Computing
Next wave
5G/6G
AR & VR
Sensor
Tech
Sustainability Tech
Space Tech
Smart Robots
Autonomous Driving
AI/ML
Blockchain
3D Printing
Zero Trust Architecture
Intrusion
Detection
Key themes:
Haptics
Innovation Maturity
(60%-# of patents,
40%-freshness of patents2
)
● Core Ops/Verticalized BPO
● Data Engineering & BI+
Visualization
● Cloud Advisory, Platform
and Services
● Digital Strategy & Services
● Digital Marketing
● IoT Applications/Industry 4.0
● Managed Security Services
and Consulting
● Engineering Services
● ERP (Enterprise Solutions)
● Managed Services for
Data Centers
Next-gen Platform
Solutions
AI-enabled
Data Solutions
IoT Systems
& Platforms
AI-enabled
Cybersecurity
AI/ML Connectivity
Blockchain Next-gen
Computing
Digital
Manufacturing
Climate-change
Services
Immersive
Media
Key technologies with high innovation maturity & funding momentum. Expected to disrupt market in 3-5 years
Top 5 identified technological disruptors - enterprise spend and market backed customer survey approach
for existing TAM

13
Source: Gartner, BCG Analysis
Note: 1. Others include Education, travel, energy and retail industries
Global AI market is expected to reach $320-380 Bn by 2027 with GenAI
expected to contribute ~33% & Financial expected to be the largest sector
Global AI market size forecast
by Technology (2023-2027)
Global AI market size forecast
by Solution (2023-2027)
($Bn)
Machine
Learning
& Vision
GenAI
Deep
Learning
NLP
2023
$110-
130 Bn
$110-
130 Bn
2027E
38%
32%
17%
13%
13%
25%
29%
33%
$320-
380 Bn
$320-
380 Bn
25-35%
CAGR
25-35%
CAGR
Global AI market size forecast by
Industry (2023-2027)
2023 2027E
% of overall
AI market
27%
30%
Financial
14%
15%
Media
9%
5%
Healthcare
8%
9%
Tech
14%
13%
Government
12%
13%
Manufacturing
17%
15%
Others1
Increasing accuracy:
With better enterprise data
engineering practices & use case-
specific model training
Computing power optimization:
While hardware OEMs are designing more
powerful AI chips & processors, model/
algorithm fine-tuners are optimizing the
need for compute power
GenAI democratizing AI for all:
With user-friendly, natural language interface,
interaction with complex Data & Analysis
backend is within reach of all, paving the way
for a future with AI in every product or service
Key trends driving market growth
Services
Hardware
Software
2023 2027E
58%
24%
18%
66%
22%
12%
AI Market Overview

14
Keeping with the trend, enterprise IT spend on AI/ML capabilities likely to
experience significant increase relative to last year
Please select the top 3 products where you expect your company to have the largest IT spend increases/decreases (in terms of %) over the next 12 months.
Source: BCG & GLG IT Buyer Pulse Check 6.0 (August 2023), N = 368, IT Buyer Pulse Check 5.0 (December 2022), N = 450
More likely to
decrease IT spend
More likely to
increase IT spend
AI/ML (general capabilities), including GenAI
Cloud services
Security infrastructure
Analytics
App dev/DevOps/Integration
ERP: Finance/accounting/treasury
CRM: Digital commerce platforms
ERP: Traditional
CRM: Customer service and support/field services
CRM: Marketing/MarTech/AdTech/Sales
ERP: HR management
IT operations management
Communication, collaboration and content management
System & service management
Network infrastructure
Storage infrastructure/data management
Devices
Server infrastructure
28%
28%
26%
26%
12%
9%
5%
3%
3%
2%
-3%
-8%
-10%
-12%
-12%
-15%
-20%
-33%
AI Market Overview

15
AI market in India projected to grow at ~25-35% CAGR till 2027, supported
by a large AI talent base and high AI investments
2nd highest installed talent base
with ~420K employees working in
AI job functions
5th
highest quantum of AI
investments received, amounting
to ~$4 Bn in 2022 & 2023
2027E
$17-22 Bn
$7-9 Bn
2023
Source: Gartner, State of Data Science & AI in India – NASSCOM Report 2023, Quid, BCG Analysis
Note: 1. Other includes Healthcare, Automotive, Retail and Travel Industries
AI market size forecast in India (2023-2027)
25-35%
CAGR
15-17%
23-25%
15-17%
14-16%
9-11%
9-11%
7-9%
Tech (Products
& Startups)
BFSI
Media
Public sector
Tech Services
Manufacturing
Others1
18-20%
13-15%
11-13%
10-12%
23-25%
8-10%
9-11%
AI Market Overview

16
Global Investments | Investments in AI at $83 Bn in 2023,
growing at ~25% CAGR
Global AI investments (2019-2023)
Top vertical & horizontal themes in AI investments1
(2022-2023)
Key trends (2022-2023)
Tech, Banking & Healthcare
attract a majority of the AI
investments
Data & Analytics, GenAI,
and Supply Chain topped
the agenda for investments
made in AI
~$83 Bn
~$35 Bn
2023
2019
+24%
CAGR
40%
64%
19%
38%
7% 6%
14%
6%
15%
20%
28%
42%
9%
Source: Quid, BCG Analysis
Note: Analysis based on all M&A, private investment and minority stake investments made in AI from 2019 – 2023; 1. Investments can span across multiple horizontal themes and
verticals; 2. Others include Robotics & RPAs, AR/VR Platforms, Contract Management & Legaltech, Digital Content, Semiconductors and Test Automation Software; 3. Media - Gaming &
Entertainment; 4. Others include Manufacturing, Energy & Sustainability, Insurance, Defense, Education, Real Estate and Agriculture
Data & Analytics
GenAI
Cybersecurity
Marketing & Advertising
solutions
Others2
HR/CRM solutions
Supply Chain
Tech
Banking
Automotive
Media3
Others4
E-Commerce & Retail
Healthcare
Horizontal
AI themes
Vertical
themes
33%
AI Investments Landscape

17
Verticals funding | Tech, Banking & Healthcare attract the most funding globally;
Data & Analytics and GenAI receive the highest across the three sectors
Note: Key deals: Investor name (Target name, deal size, use case); Analysis based on all M&A, private investment and minority stake investments made in AI from 2022–2023; The list of investors and deals mentioned is not exhaustive. ;
1. Media - Gaming & Entertainment; 2. Investments can span across multiple verticals, E.g., OpenAI is a Tech company with use cases in AI for multiple industries like Banking, Healthcare, E-commerce, Media, etc.
Tech (64% funding)2
Key AI themes & use cases
● Data & Analytics: Analytics automation
● GenAI: Speech recognition, Prompt
text generation
Key deals
Microsoft (Nuance Communications,
~$20 Bn, Speech recognition),
Microsoft (Open AI, ~$10 Bn, Prompt
text generation)
Banking (38% funding)2
● Data & Analytics: Market intelligence,
Working capital management platform
● GenAI: Chatbots, Virtual
assistants, Fraud detection
Key deals
SAP (Taulia LLC, ~$780 Mn, Working
capital management), Goldman Sachs
& Alphabet Inc. (AlphaSense, ~$325 Mn,
Financial market intelligence)
Healthcare (33% funding)2
● Data & Analytics: Patient health
analytics/prediction
● GenAI: Small molecule drug discovery
Key deals
Kairos HQ (Cera Care, ~$320 Mn, Patient
analytics), Andreessen Horowitz & Nvidia
Ventures (Genesis therapeutics, ~200 Mn,
Drug discovery)
Automotive (15% funding)2
● GenAI - Personal retail assistance by
capturing customer's cues (visual & text)
● Data & Analytics – End to end
E-Commerce channel optimization &
automated analytics
Key deals
MasterCard (Dynamic Yield, ~$320 Mn,
Personal Retail Assistance), Insight Partners
& Trinity Ventures (CommerceIQ, ~$115 Mn,
E-Commerce channel optimization)
● Data & Analytics – Location intelligence,
Autonomous vehicles
● Supply Chain – Integrated global supply
chain systems
Key deals
Thoma Bravo (Nearmap, ~$750 Mn,
Location intelligence), Carlyle Group &
Robert Bosch ( JingChi Inc., ~$400 Mn,
Autonomous vehicle)
Media1
(14%)2
Data & Analytics, Marketing & Advertising
solutions
Manufacturing (14%)2
Data & Analytics, Supply Chain,
HR/CRM solutions
Energy & Sustainability (10%)2
Supply Chain & waste management,
Data & Analytics
Defense (5%)2
Data & Analytics, Cybersecurity
Others
E-Commerce & Retail (20% funding)2

18
While Data & Analytics emerges
as the key theme across all
regions, GenAI also emerges as a
key theme in NAMR
While Data & Analytics emerges as
the key theme across all regions,
HR/CRM solutions also emerge as
a key theme in Europe
While Data & Analytics emerges
as the key theme across all
regions, Marketing & Advertising
solutions also emerge as a key
theme in APAC
Regional funding | North America emerges as the leading investment hub
for AI; Data & Analytics receives the most funding across regions
North America Europe APAC
68%
~$116 Bn
11%
~$20 Bn
15%
~$26 Bn
Funding
4
5 Marketing &
Advertising
solutions
Cybersecurity
1 Data &
Analytics
5
2 GenAI
GenAI
3
3 Supply Chain
3 Supply Chain
Supply Chain
1 4
Data &
Analytics
Contract
management
& Legaltech
2 HR/CRM
solutions
1 4
5
Data &
Analytics Test
Automation
Software
2 Marketing &
Advertising
solutions
Semiconductor
Top AI
use cases
Note: Analysis based on all M&A, private investment and minority stake investments made in AI from 2022–2023.
Remarks
1% Canada 25% Germany 3% Spain
7% Japan
16% India
47% United Kingdom
99% United States
10% France
6% Singapore
52% China
7% Australia
4% Switzerland
Top
countries

19
Note: Analysis based on all M&A, private investment and minority stake investments made in AI from 2022–2023; Media - Gaming & entertainment, Energy - Energy & sustainability
Data & Analytics strong across regions and verticals; GenAI
skewed to NAMR; High Marketing traction in India
Tech
Banking
Healthcare
Automotive
Media
E-comm.
&
Retail
Manufacturing
Energy
Defense
Insurance
AI use cases/
Industries Regions
1
2
3
4
Cybersecurity
Data & Analytics emerges
as a key focus area
across regions especially
in Tech, Banking and
Healthcare firms
1
2 GenAI investments heavily
skewed to NAMR
Example: Microsoft –
Open AI
India seeing early traction
with Sarvam.ai launching
enterprise-grade GenAI
platforms
3 HR/CRM solutions emerge
as a Europe-specific bet
especially in Tech, Banking
& Healthcare firms
Example: Keyrus offers
enterprise performance
management services
4 Marketing & Advertising
solutions emerges as a key
focus area in India across
verticals
Example: Rubick.ai offers
AI based E-commerce
cataloging solutions
Cybersecurity gains
moderate traction across
Defense, Tech & Banking
verticals
Example: AI processes
large amounts of data
for defense & military
organizations to detect
security threats
5
NAMR
NAMR
NAMR
NAMR
NAMR
NAMR
Europe
Europe
Europe
Europe
Europe
Europe
APAC
APAC
APAC
APAC
APAC
APAC
India
India
India
India
India
India
Data &
Analytics
GenAI
Supply Chain
HR/CRM
Solutions
Marketing &
Advertising
solutions
5

20
Key AI horizontal themes & verticals for investments made by Indian Tech Product & Services1 investors
~55% of investments made by Indian Tech Product & Services investors are
made in India with focus on Data & Analytics, Digital Content and GenAI
Investments ($Mn) made by Indian Tech Product & Services investors (2022-2023)
Major deals
Jio Platforms invested
significantly in Glance
Digital to enable
personalized content via
AI on Jio phones'
lock screens
Perfios acquired Karza
Technologies to build
a one-stop-shop by
leveraging Karza's
expertise in fraud
prevention through
superior data engineering
Infosys acquired BASE
Life Science to aid global
pharmaceutical companies
in accurately analyzing data
related to clinical trials,
drug discovery, patient's
health, etc. using AI
Data & Analytics 29%
27%
25%
10% 10%
Digital Content
Marketing & Advertising solutions
GenAI
Tech 55%
15%
11%
Healthcare
Media3
Banking
Key horizontal AI themes Key verticals
Note: Analysis based on all M&A, private investment and minority stake investments made in AI from 2022–2023; 1. Tech Product & Services firms include Application software, Systems software, Data processing, Hardware, IT services and
consulting firms ; 2. $ value for 2/26 deals undisclosed; Media - Gaming & Entertainment sector
~$730 Mn2
~68% of investments made by Indian Tech investors are
done via M&A v/s ~75% for global Tech investors

21
2022
2027E
Demand for AI talent in India expected to grow at 15% CAGR till 2027 to
serve the AI market
Source: BCG Analysis, 1. NASSCOM State of Data Science & AI talent in India Report - Feb'23 2. HAI AI Index Report 2023 3. LinkedIn - Future of work Report
India ranks among the
top 5 nations with a 14x
growth3
in individuals
skilled with AI in the
last 7 years
AI talent demand in India (no. of people, '000s) Growth in #individuals skilled with AI
(Top 5 countries)
600-650
1250-1350
15%
CAGR1
20x
Singapore
15x
Ireland
13x
Canada
16x
Finland
14x
India
India has the highest
skills penetration with
~3x2
more AI skilled
talent than
other countries
Top 5 countries' AI skills penetration
Canada
Israel
Germany
India 3.23
2.23
1.72
1.65
1.54
United States
Perspectives on Indian AI Talent

22
Source: LinkedIn talent insights, BCG Analysis
Note: Analysis done on data captured at end of Dec’23 v/s Dec’22, # Tech companies analyzed=21; 1. AI/ML job titles include AI/ML engineers, consultants, managers, researchers, specialists, directors, VPs 2. Large Tech firms– revenue >$1
Bn, Mid Tech firms–revenue $200 Mn - $1 Bn 3. Offshore jobs – Jobs in India
Already seeing 15%+ growth in AI/ML jobs1
in India in the last 12 months
with positions like AI Engineers growing at 70%+ YoY
Growth in #AI/ML job titles for Tech & BPO firms from 2022-2023
Hiring trends in AI/ML job titles1
BPO
xx % - 1 year growth rate in #employees
15% 18% 6%
0
0
60
60
100
100 150
150 250
250
Entry & Mid AI/ML job titles
Senior AI/ML leadership job titles
Artificial Intelligence Engineer
AI/ML Consultant
AI Manager
AI/ML Researcher
Senior ML Engineer
AI/ML Specialist
AI Vice President
AI Director
Head of AI
11% 8% 12%
Senior
AI/ML
leadership
Large
Tech
Mid
Tech
Total AI/ML
jobs
Dec 2022 Dec 2023
xx% YoY Growth
67%
14%
32%
11%
14%
9%
14%
7%
6%

23
Note: Analysis done on data captured at end of Dec’23 for Tech services & Tech product companies, # Tech comapnies analyzed = 21; 1. Large Tech firms– revenue >$1 Bn, Mid Tech
firms – revenue $200 Mn - $1 Bn
ML, Python & SQL dominate current skills requirement in
AI; While a few GenAI skills gradually emerge
C1 C2 C5 C8
C3 C6 C9
C4 C7 C10 C11 C14
C12 C15
C13 C17 C18
C16 C19 C21
C20
Data & Analysis
Data Science
Business
Analysis
Software
Development
Agile
Methodologies
Machine
Learning
Skills/
Companies
C /C++
Python
Java Script
SQL
Java
Cloud
Computing
AWS
ML, Python & SQL
emerge as the top
3 skills across most
companies
1
BPOs have a more
analytically driven &
skilled installed talent
than Tech firms
2
AI employees in Large
Tech Services firms
starting to pick up skills
in Github, PyTorch &
Databricks more actively
3
Programming
Large Tech Mid Tech BPOs
Software
development
Analytics
Cloud
1
Key skills possessed by employees
2
Emerging
GenAI
skills
GitHub PyTorch Databricks

24
Americas
13%
EMEA
12%
APAC
75%
Concentration of AI talent by geographies Top 5 countries for AI talent
While Large Tech and BPO are hiring heavily offshore, Mid Tech are
focusing on offshore + nearshore
BPOs
India
69%
USA
11%
UK
2%
Phillipines
2% 2%
Colombia
Mid Tech
India
42%
USA
10%
UK
4%
Ukraine
6%
Poland
5%
Large Tech
India
75%
USA
9%
UK
3%
Canada
2%
France
2%
Note: Analysis done on data captured at end of Dec'23 v/s Dec'22 for Tech services & Tech product companies; # Tech companies analyzed = 21; 1. Large Tech firms– revenue >$1 B, Mid Tech firms – revenue $200 Mn - $1 Bn

25

26
02
AI Acceleration Framework & Imperatives
for Tech Services

27
Technology
Digital platforms
Models
Cloud & Infra
Security Policy
Data
Data storage
& quality
Data labeling &
Metadata
Data governance
Data encryption
Op Model
Partnership
ecosystem
Use case
prioritization &
financing
CoE & organization
model
Enterprise agility
People
Workforce planning
Talent readiness
Impact on roles
and culture
Training &
upskilling
Comprehensive multi-dimensional AI Acceleration Index Framework used to
assess AI maturity in the Tech services industry
Governance
framework
Strategy
Vision & Governance
AI for Clients
Use case
scalability
Client
Readiness
Key offerings &
Service maturity
GTM strategy
Value
proposition
AI leadership
Ethical AI
policies
Roadmap &
success metrics
C-Suite alignment
& vision
X-functional
stakeholders

28
Detailed understanding of 65+ players
across the framework
Approach | AI Survey question
example: Leadership buy-in & vision
Overall leadership alignment to the AI & GenAI vision – essential to drive the
company's AI agenda as well as establish strong governance practices
How strong is the senior leadership’s buy-in for AI (including GenAI)
initiatives, and how clear and transformative is the organization’s vision
for integrating AI (including GenAI) into its operations?
Equitable spread across Large and Mid Tech,
BPO & GCC companies
Insights from C-suite respondents across large players
Senior leadership shows limited
understanding and commitment to AI
0
Some senior leadership buy-in & vision
emerging, not fully defined/communicated
33
Senior leadership committed to
AI initiatives; Vision aligned
66
Strong senior leadership buy-in and
clear AI vision
100
30%
24%
20%
26% Mid Tech
Large Tech
SVP, VP
C - Level
GCC
BPO
58%
42%
Source: NASSCOM-BCG Tech Services Industry AI Maturity Assessment Survey, BCG Analysis
AI maturity
score
Answer options

29
Key Highlights across the framework
● 30% of services players have matured to offer advanced AI & GenAI services (e.g., data engineering, AI model finetuning) and are
leveraging it to build a growing AI portfolio of business (15% of services firms expect to add 20+ clients in the next year)
● While 60% of those surveyed still use finetuned LLMs for most use cases, others are in the process of developing customized
SLMs or LLMs for specific use cases
● Services players have also demonstrated a capability to put client use cases to production; 10% of players have 10+ GenAI use
cases in production vs industry average of 4-5
● Services players are using soft launches to gauge & increase client interest; they have been able to leverage unified GenAI
platforms to accelerate time to market for newer PoCs from months to weeks
● 70%+ services players already have a dedicated & specialist AI head to drive AI initiatives while 55%+ also have a central CoE
with cross-functional governance forums with business, tech & legal participation
● 70% services players have clearly defined metrics to monitor the progress of AI initiatives and have established a well-defined
prioritization framework for directing AI efforts across business impact x feasibility x ability to replicate & scale
● Most of the firms have started to think about their ethical AI policies and have a draft in place, but only 18% have comprehensive
ethical AI policies with a dedicated senior ethical AI compliance role
● Training & upskilling continue to be the preferential route to attract the right AI talent. However, AI roles & leadership are also
seeing accelerated hiring with c. 70% increase in AI engineer and c. 15% increase in key AI leadership roles
● 70%+ companies have AI partnerships with CSPs; some are also partnering with niche startups for differentiated capabilities
● Change management across cultural resistance to the adoption of GenAI in delivery as well as the adoption of ethical AI
practices remain the key imperatives for most tech services firms
Execution
against Client
agenda
Future ready
organization
roadmap
Internal
Capability
build & GenAI
adoption
1
2
5
6
7
8
9
10
3
4

30
Note: 1. Average score for Large and Mid technology players & BPOs
AI for Clients1
63
Vision & Governance
67
Operating Model
61
People
57
Technology
55
Data
69
Summary overview of the 6 dimensions
Average score of Large & Mid
Technology Players, BPOs & GCCs

31
AI for Clients | Overview
0 10 20 30 40 50 60 70 80 90 100
Average score
(out of 100)
Tech services' maturity
& range of offerings
A
Scalability of AI offerings
from Pilot to enterprise-
wide deployment
B
Differentiation of
offerings vs peers
C
D
AI GTM strategy
across verticals
Client appetite &
readiness for AI projects
E
66
Technology Players & BPOs
xx
66
57
58
61

32
30% of firms surveyed offer full stack of services (advisory
to model finetuning) to help build a solid AI portfolio
>20 clients
11-20 clients
6-10 clients
<5 clients
Basic advisory
services
+App development
& maintenance
+Data engineering
+AI model building
(finetuning) services
AI portfolio growth
(#clients expected to be onboarded in
the next 12 months)
Tech services service maturity &
range of offerings
15%
30%
20%
5%
35%
18%
33%
45%
To make sure we have end-
to-end control on infra costs,
we are offering a GenAI
platform as a service for our
clients which enables them
to fine tune & test models,
as well as develop apps for
different use cases
Mid Tech player
AI for Clients

33
Robust use case prioritization & financing frameworks enable higher % of
POCs in production, with track record of successful at scale deployment
Scalability of offerings
25%
50%
25%
0%
100
0
Robust approach to scaling AI
initiatives; Track record of successful
enterprise deployment
Defined strategies for scaling AI; Some
pilots have successfully scaled
Initial scaling strategies in place;
Challenges in enterprise-wide deployment
No clear strategy for scaling AI; Pilots
rarely progress to deployment
We use internally developed frameworks which evaluate ROI,
Productivity improvements, internal CSAT & Scalability amongst
other factors before clearing a use case for client
We've engaged with two-thirds of our client base on AI &
GenAI work, ~30 proposals on-going and have 6+ use cases
in production
BPO player Large Tech player
AI for Clients
<= 10 use
cases
11-15 use
cases
16-20 use
cases
>20 use
cases
40% 20% 13%
28%
GenAI use cases identified
None <= 5
cases
6-10
cases
>10
cases
8% 43% 10%
40%
GenAI use cases in production & scaled

34
Case Study | AI
& GenAI projects
deployed in
production
Multiple use cases in
production, including:
● Real time speech to
speech translator
CoPilot between Japan
& US teams for reduced
manual translator &
efforts for an
insurance company
● Angular to React.js
code conversion for a
Big 4 company
● Customized fraud
detection solution for a
large insurance player
Large
Tech firm
Multiple use
cases by BPO
An Exemplar BPO has
multiple use cases in
production, including:
● Fraud detection for an
insurance firm by deploying
AI algos at every decision
point of the value chain
(adjudication, payment
calculation, litigation, etc).
• Used GenAI to classify
degree of damage
based on the claims
submitted via email
● Build accurate responses
to customer queries using
GenAI to process multiple
checkpoints for a
travel intermediary
For a hospitality & hotel
chain, a Mid Tech player
worked on a document
rationalization project
● GenAI CoE and
incubator created
● Identified 50+ use cases
● Standardized 40K JDs to
120 roles by processing
and ensuring task to role
mapping using AI
Mid
Tech firm
AI for Clients

35
Exemplars have holistic x-vertical GTM strategies focused
on farming logos; Some sell outcome-based projects ● Focusing on farming
existing logos with AI &
GenAI pilots
● Prioritizing easily replicable
use cases
● Differentiating themselves
via POCs, demonstrating
accuracy at scale, data
& ethical AI policies, &
integration with client
environment allowing them
to showcase X-functional
expertise
● “Soft launches” seen as
a path by Tech leaders
to demonstrate tangible
value to clients within a
short span of time & secure
larger/full scope AI projects.
E.g., a Tech player offers
clients pilot solutions with
defined KPIs (Productivity,
Quality & Time to Market)
Clear comprehensive plan across all verticals
Well-defined AI GTM strategy for most verticals
Beginning to define AI GTM strategy for some verticals
50% companies with
GTM defined for
75-100% verticals
No clear AI GTM strategy for any vertical
23%
28%
50%
0%
Depth of GTM strategy across verticals
We have a crossfunctional (across BUs and CoE) team to look at
GenAI GTM & offering focus. Each BU has thought carefully about
which subsector(s) to target for maximum initial traction on GenAI
use cases (E.g., P&C Insurance in BFSI) as well as has been given
resources & investments to run pilot projects on prioritized offerings
We pitch AI/GenAI driven outcome-based projects directly to the
BU head instead of an AI head, showcasing our domain expertise
& ability to deliver outcomes
Mid Tech player
BPO player
AI for Clients
GTM strategy for most players:

36
To address challenges around clients maturity, awareness & ability to invest,
Exemplars deploy client initiatives such as co-sponsored hackathons
While GenAI has generated large interest across clients, limitations observed in clients':
● Awareness of AI applicability for use cases & ability to invest
● Maturity of Tech infrastructure & data policies
● AI workforce maturity and leadership buy-in
Exemplars have been able to prioritize & pursue after the right set of existing logos who have high willingness & capability to adopt AI E.g., Exemplar Mid
Tech ran a survey among top logos to understand & prioritize, basis willingness to adopt GenAI
Exemplars also invest in coaching & raising awareness of AI/GenAI capabilities amongst their clients. E.g., a BPO is building AI literacy amongst clients
through co-sponsored hackathons
Fully equipped to engage
with AI; Specialized teams,
robust strategies etc.
100
0
Distribution
Moderate preparedness,
some dedicated resources;
Undertaken few AI projects
Early interest in AI
projects; Lacking structured
approach & resources
Client teams, organizations,
industry not prepared to
undertake AI projects
18%
50%
33%
Client readiness for AI
AI for Clients

37
Vision & Governance | Overview
Average score
(out of 100)
Leadership vision
& buy-in
A
Dedicated leader to
oversee GenAI efforts
B
E
Comprehensive AI
Governance framework
and SOPs
Stakeholder involvement
& engagement in
value assessment
C
Clearly defined
organization roadmap
for AI initiatives
D
Ethical AI policies and
compliance
with regulations
G
Clearly defined success
metrics for AI initiatives
F
70
67
64
56
62
82
69
0 10 20 30 40 50 60 70 80 90 100
xx

38
Exemplars have hired dedicated & specialist AI leaders with strong credentials
and are driving top-down implementation of the AI vision
Dedicated AI Leadership
In-process of
identifying a leader
14%
Leader identified,
yet to take charge
14%
Dedicated executive
leading AI initiatives
68%
No dedicated person
4%
Senior leadership
committed to
AI initiatives;
vision aligned
58%
Strong senior
leadership buy-in
and clear AI vision
26%
Some senior
leadership buy-in
& vision emerging,
not fully defined/
communicated
16%
Senior leadership buy-in & clear vision to integrate AI into operations
● In the last ~2 years, several dedicated ‘AI Head’ positions
have been recruited externally
● Preference to recruit externally for leadership roles to
get a jumpstart on AI capability building
● ~15% growth witnessed in key AI leadership roles across
Tech and BPO firms
We are looking to hire specialist AI leaders who not only possess
deep understanding of technology but also business acumen &
experience to execute the vision. Today, we have also been able
to execute against our AI agenda across BU's and clients to drive
transformation thanks to the steep ramp up & cultural shift driven
by the joint ownership over the AI agenda by our AI &
CXO leadership
We believe in training the whole organization in AI including our
board members and not just domain-specific employees
BPO Exemplar
BPO Exemplar
Vision & Governance

39
Vision & Governance
Bespoke AI governance frameworks with oversight from x-functional
stakeholders to ensure holistic prioritization and value
Comprehensive AI Governance
Framework and SOPs
Involvement of x-functional stakeholders in
value assessment of AI initiatives
Exemplars have established dedicated AI steering committees
which convene at frequent intervals on topics regarding:
● Funding & investments
● Security/IP/regulatory compliances
● Result monitoring of Pilot use cases
● Use case prioritization for production across clients
● Assess commercial outcomes of client offerings
Effective committees have a good mix of technical (CTOs, AI/
GenAI heads etc.), business (BU Heads), and legal expertise
Exemplars have also created “AI/ML Governance Specialist”
roles to enforce data accessibility, quality and
regulatory compliance
52%
20%
24%
4%
30%
46%
20%
4%
We used to have a quarterly steering committee meeting on AI
earlier. Now we meet on a bi-weekly basis to discuss learnings
& chart out the roadmap
Mid Tech player
Involvement of cross-functional stakeholders in value
assessment ensures diverse and inclusive perspectives. It's
integral to defining and assessing the value of GenAI projects
Large Tech player
Basic governance framework
& SOP for AI exists
Robust, adaptive AI
governance framework & SOP
Defined AI governance
framework & SOP that is
regularly reviewed
Lacks AI governance
framework & SOP
Some involvement in
value assessment process
Minimal involvement
Regular involvement in
value assessment; Scope
for deeper involvement
Comprehensive x-functional
stakeholders involvement
in value assessment

40
Case Study | GCC of a Global Bank has a detailed x-functional AI & GenAI
governance model
● Core function is to execute
against AI vision & roadmap;
Implement & regularly track
metrics and conduct
reviews regularly
● CoE works with the CIOs,
BU Heads, Data Officers to
experiment – creates a x-BU test
bed environment
● CSO monitors compliance of
guardrails of the data usage
by the CoE
Core activities include defining:
● Frameworks on use case financing
& prioritization on GenAI
● Data security, ownership, usage
norms, IP norms including
guardrails on data usage,
storage, etc.
● Compliance/risk &
regulatory frameworks
● Ethical AI frameworks & review
cadence bi-weekly
Chaired by:
● Vision: Defines centrally steered
vision, goals & ambitions for the
AI agenda globally
● Roadmap: Sets quarterly/annual
& 5-year roadmap across people,
data, tech, op model, governance
& client offerings
● Quarterly review
of implementation
CIO: Chief Information Officer CSO: Chief Security Officer CoE: Center of Excellence
Stakeholders
Oversight by Global ExCo Governance Forum Execution & Monitoring
Chief Operating Officer
Chief Digital Officer
All CIOs of the group
All BU heads of the group
All Data Officers of the group
AI CoE
CSO of each BU
Responsibilities
Vision & Governance

41
Clearly defined roadmaps and success metrics to monitor client outcomes
for value generation and associated risks
Players are doing well on defining an AI roadmap &
success metrics Exemplars have built comprehensive quarterly & annual roadmaps primarily
focused on building both internal & external AI capabilities & infrastructure
As part of the roadmap, Exemplars have also clearly defined AI success metrics/
evaluation frameworks (primarily developed internally) monitored regularly by a
dedicated AI team
Defining AI roadmap
Success metrics
Comprehensive,
forward- looking,
evolving roadmap
26%
Defined roadmap
with periodic
AI updates
56%
Basic roadmap,
lacks detail/
strategic focus
18%
Comprehensive
success metrics,
driving AI performance
evaluation
24%
Clearly defined
success metrics for
key AI initiatives
46%
Basic success
metrics in place
30%
Exemplar
Has built BU wise metrics along
the axes of
BPO Bank's GCC
Efficiency: Productivity
improvements, E.g., cost or
manpower reduction
Effectiveness: Improved
agility/responsiveness
of processes E.g.,
TAT reduction
Experience: Measured
impact of user experience
through NPS CSAT
Business facing
● CSAT
● Bank's revenue
● ROI
Internal
● Risk & regulatory
● Cost outlay
Vision & Governance

42
Majority players are early in their Ethical AI journey with focus on basics
like bias detection
Ethical AI data practices being applied at 2 stages:
● While creating datasets: leverage analytics to ensure that sufficient diversity is
built into the LLM
● While utilizing datasets: LLM parameters are defined to filter out any biases
E.g., if there's a gender related question creeping in, then you can select / filter out
these questions
E.g., a BPO player has built guardrails that disallow the identification of demographic
characteristics while dealing with job applicants' datasets for its US clients
Exemplar | 4-dimensional Responsible AI framework built by a Large Tech player
Individual dimension Social dimension Environmental dimension
Responsible data handling in
preparing the right datasets
with equality and equity
in mind
AI impact assessments on
people and communities
Data minimization, smart
data processing approaches,
synthetic data to limit
data extraction
Protection against attacks
that undermine privacy,
pollute outcomes, lead to
unfairness/discrimination
Technical dimension
While Ethical AI is a pre-requisite for all clients, it is especially important while
dealing with clients in highly regulated sectors like Banking, Insurance & Healthcare.
Also essential to build in geographic regulation specific nuances
Mid Tech player
Maturity of Ethical AI policies
18%
42%
32%
8%
Leader in
ethical AI
Clear ethical
AI policies
Some ethical
AI policies
No clear
ethical AI
policies
Vision & Governance

43
Operating Model | Overview
Average score
(out of 100)
A
Organization model and
CoE setup
B
Clearly defined BU & CoE
relationship & ways
of working
C
Integrated approach
to financing use cases
across BU, SL, & CoE
E
Use case prioritization
and planning
D
Leveraging external
partnerships for
competitive AI positioning
58
57
62
58
69
0 10 20 30 40 50 60 70 80 90 100
xx

44
Exemplars have set up large AI & GenAI CoEs with seamless ways of working
with the BU's; Some have 500+ large CoE constructs
Set-up of CoE
30% 12%
26% 32%
Well-established CoE No CoE for AI
efforts
Partially established CoE In-process of establishing a CoE
Typical functions carried out by CoE
Prioritization of
use cases
Monitoring of AI
initiatives
Budgeting & financing
of use cases
Monitoring outcomes
of AI client offerings
Driving external
partnerships & consultancy
50%+ companies
have at least a
partially established
CoE - which is an
important function
to take the AI
agenda forward
Exemplars have
well-established
CoEs with dedicated
workforce:
Another Large Tech exemplar has
established a GenAI & LLM CoE
with 1.6K professionals with a
commitment to invest $3 Bn in the
next 3 years towards its Data and
AI practice
Exemplars also focus on having a
well-defined hub & spoke model
with dedicated representatives in
each BU for well integrated ways of
working between the CoE and BU:
Our AI CoE has dedicated
SPOCs to liason with each BU
to ensure integrated ways
of working
We have a 500-600 member CoE focused
on AI & GenAI. Over the last 2-3 years
we have built 2 stellar offerings that
serve several Fortune 500 companies.
Additionally, resources from this CoE
assist for short stints on AI requirements
across multiple organization
wide projects
Mid Tech player
Large Tech player
Operating Model

45
Exemplars have established frameworks prioritizing use cases based on
business impact, implementation feasibility and ability to scale & replicate
An Exemplar prioritized
pushing a fraud analytics
solution for an existing banking
client to production as it could
then also be used for warranty/
coupon management for a new
logo in the consumer durables
space as a pilot
Exemplars tend to have a seamless budgeting framework between BU, SL, and Central AI organization, ensuring
optimal allocation of resources, clear financial accountability, and efficient fund utilization
Exemplars do exceptionally well on prioritization of use cases by assessing:
Business impact
financial attractiveness, digital
adoption, differentiated
value proposition
Feasibility
basis current experience,
implementation feasibility, and
overall risks associated
Other factors
such as replicability,
effort to scale, etc.
Planning & prioritization of use cases
Sophisticated prioritization
framework fully embedded in
strategic planning process
Lack of formal prioritization
framework for AI use cases
Well-defined prioritization
framework in place with AI use
cases ranked effectively
Prioritization framework in
development with some criteria
established for ranking AI
use cases
30%
52%
14%
4%
Financing of use cases
Seamless budgeting framework
for AI projects
Unclear budgeting for
AI projects
Budgeting for AI projects is
well-structured
Basic framework for AI
project budgeting
20%
54%
20%
6%
Operating Model

46
Financial attractiveness
(ROI, attractiveness wrt
market potential & growth,
market mindshare)
Digital adoption
(extent of efficiency
improvement and cost
reductions)
Differentiated value
proposition
Case Study | Use cases bucketed into 5 categories based on
a firms' in-house business impact x feasibility framework
Clear criteria for prioritizing the long list of productized offerings
Other factors: Strategy & vision alignment, SBU head buy-in, replicability, effort to scale, compliance risks
Quantitative parameters Qualitative parameters
Business Impact
Feasibility
Star performers:
Invest heavily & develop
significant capabilities
1
New lucrative bets:
Quickly ramp-up
capability and
innovate solutions
3
Re-assess:
Whether the offerings
make business sense
4

De-prioritize:
Limit further
investments in low-
return areas
5
Solid bets:
Continue to invest and
develop capabilities
2
Current experience
& depth
Implementation feasibility
(Talent availability & skills, IP)
Risks associated
(Regulatory, business)
High
Medium
Low
High
Medium
Low
1
2
3
4 5
Operating Model

47
Players have strong partnerships, showing early success in pilot deployment;
Exemplars effectively leverage partnerships to reduce cost & time to market
Building capabilities from scratch often seen as a barrier. Best in class
mid-size players work with partners to get solutions into place quickly
A BPO player has launched an AI platform in collaboration with
Google AI, IBM Watson and Microsoft Azure
Popular external partnerships across Tech Service players
An Exemplar has observed ~20% productivity increase by leveraging
Copilot GenAI solution in pilot - after GitHub Copilot introduction,
developers' submissions increased by 44% and implemented an HTTP
server in JavaScript, which was 56% faster than baseline
In addition to partnerships with CSPs, firms are partnering with niche
firms for domain specific models.
E.g., Large Tech player has partnered with Writer for knowledge
enhancement capabilities
We have a partnership with Microsoft for Azure, Co-Pilot, etc. For
LLMs, we either buy from open market or develop in-house. Even
for hardware we have entered into 3rd party partnerships
Leveraging external partnerships
BPO Exemplar
Established partnerships,not
fully integrated
in talent strategy
Adhoc engagements
with external
partners
Strategic & effective
use
of partnerships
for
competitive advantage
No engagement
with
external partners
GitHub
Copilot
Amazon Code
Whisperer
Microsoft
Azure
48%
24%
12%
16%
Operating Model

48
People | Overview
0 10 20 30 40 50 60 70 80 90 100
Average score
(out of 100)
Impact of AI on job roles
& workflows
A
Resistance management
& cultural shifts for
AI adoption
B
Workforce planning to
secure necessary AI skills
& support
C
Training & upskilling
employees to work
with AI
D
Share of delivery
workforce trained in AI
E
65
34
58
66
62
xx

49
Players assessing the impact of GenAI on workflows and expect it to have a
positive impact on productivity; Crucial to plan for resistance management
Impact analysis of AI on
job roles & workflows
Exemplars have found that the expected impact of AI & GenAI has potential to positively impact Software Engineers
& Product Managers across:
In-depth impact
analyses and
established new
roles for AI
Detailed impact
analyses and
defined some new
roles for AI
Initial impact
analyses conducted
(not comprehensive)
30%
34%
36%
● Higher productivity across all key roles (E.g., product managers can leverage no-code/low code solutions
to unlock time saving)
● Enhanced quality of work through capabilities like task automation, intelligent insights, error
identification, etc.
● Accelerated upskilling for junior and newly hired engineers
Exemplars have run pilots
leveraging CoPilot with their
Software Engineers – early results
have shown a 60-75% increased
satisfaction and well-being
AI & GenAI have the potential to demotivate employees & make
them fearful for their jobs. Crucial to engage with workforce
through discussions, training them on basics of GenAI, hackathons
& other investments in their upskilling
Mid Tech player
Resistance management & cultural shifts for AI adoption
No clear
strategy to
facilitate
cultural
shift
14% 48% 36% 2%
Comprehensive
strategies to
manage &
address AI
resistance
Established strategies to manage resistance and
encourage AI adoption
Few strategies to manage
resistance, early-stage
awareness for cultural shift
People

50
Exemplars introduce AI specific roles and plan for upskilling of workforce
since they recognize human capital's importance in the AI journey
Workforce planning to secure
necessary AI skills & support
Well-established,
forward-thinking
workforce planning
strategy including
AI upskilling
Pro-active approach
to workforce
planning; A few
long-term support
mechanisms
are in place
Limited workforce
planning efforts for
AI; No long-term
strategy in place
18%
18%
64%
Illustrative list of new roles
We are increasingly seeing the need
to redefine roles and responsibilities
across the organization to adapt to
new needs
We have created a Data & AI
board, made AI governance
specific organization changes &
introduced new CSO roles
GCC player
Many new roles are likely
to emerge to expand use
of AI responsibly across
the organization - the
key lies in (i) proactively
hiring the right skillset;
(ii) investing in developing
capabilities that are
difficult to hire in-house
Increased importance of
executive oversight due
to the incorporation
of AI into
business operations
Increasing complexity
& diversity of AI
applications requires
expertise to define
optimal architecture
As LLMs mature,
protection of sensitive
information becomes
a matter of
utmost importance
GenAI generates
content and makes
decisions without
human intervention
Chief AI
Officer
AI Architect AI/ML Governance
Specialist
AI Ethics &
Compliance Officer
Large Tech player
People

51
GCC player BPO player
Exemplars highlight a need for heavy initial investments in AI training
programs; Internal upskilling seen as the primary route to meet talent needs
Training & upskilling employees
to work with AI
Comprehensive, state-
of-the-art AI training
programs integrated
into professional
development paths
Good range of AI
training programs
offered; Focus on
continuous learning
& professional
development
Training programs
exist but limited
in scope; Not fully
integrated into
overall professional
development strategy
Lacks structured
training programs
for AI
44%
2%
32%
22%
There are no real GenAI experts in the market.
One can only find data scientists/engineers with
coding skills which need to be trained. Hence we
have setup an institute for AI/ML education of
our employees. This helps us upskill employees
and organically develop in-house talent
What has worked for us is hiring engineers and
data scientists – both experienced & freshers
and then training them. We have created a
3-4 months citizen data scientist program
which 300-400 people have taken over the years
% delivery workforce trained in AI
A Mid Tech exemplar has trained over 13K employees in AI & GenAI through a common platform which
provides access to:
Curated L&D material
Learnings from use cases
Reusable GenAI assets
<10%
36%
11-25%
38%
26-50%
14%
>50%
12%
Outputs for hackathons
Industry best practices on AI/GenAI
People

52
Technology | Overview
0 10 20 30 40 50 60 70 80 90 100
Average score
(out of 100)
Unified GenAI platform
A
Adoption of
unified platform
B
LLM customization
maturity
C
Security policy for AI
D
Cloud Architecture
E
44
32
46
67
80
xx

53
Players are building unified GenAI platforms but have seen mixed adoption
rates so far
Exemplar
Is there a unified GenAI platform in pilot or
production to develop use cases?
Q
~55% of respondents have a unified GenAI
platform to develop various use cases…
No existence
Design completed,
Implementation pending
Some components
implemented
All components built
& ready to use
What is the adoption of the unified GenAI
platform across use cases?
Q
8%
18% 36%
38%
… however, only ~25% of respondents have adopted the
platform for internal & client use cases
No platform/use cases
Few use cases built
All internal use cases built
Both internal & client
use cases built
A Mid Tech player launched
a GenAI platform for clients
to accelerate its ideation-to-
value journey by providing:
● Access to LLM partnerships & guidance to choose the right model
● Domain specific solutions & product offerings
● Enhanced productivity, secured guardrails & minimized bias
● Accelerated creation of new use cases from months to weeks
● Inbuilt data ingestion methodologies
8%
46%
30%
16%
Technology

54 Source: NASSCOM-BCG Tech Services Industry AI Maturity Assessment Survey, BCG Analysis
Higher Focus on finetuning LLMs or leveraging SLMs for client needs; Few
in process of developing proprietary LLMs for clients
Types of LLMs being used
Proprietary LLM fully integrated
into AI product development;
Leveraging advanced, in-house
AI capabilities
In process of developing/developed
customized LLM solutions tailored
to specific needs; Not fully
integrated into all AI products
Publicly available LLMs used
for AI product development;
Leveraging external AI models
Not yet integrated or
experimented with any LLMs in
AI product development
38%
52%
6%
4%
Exemplars
Pre-training of models (finetuning, in-context learning,
etc.), allowing them to be adapted for specific tasks
Leverage SLMs where needed, E.g., where there's a
narrow scope like in HR or Legal
Technology
Mid Tech Player
Large Tech Player
Experimented with training LLMs for different verticals
distinctively as they believe narrow LLMs lead to greater accuracy
for business cases
Finetuned the available public LLM using QLoRa technique to
adapt to client needs

55
Technology
Security Policy for AI
Exemplars have incorporated AI specific security measures
28% 46% 26%
Information security policy is fully
comprehensive and forward-thinking
with robust provisions that anticipate
and address the distinct needs of
AI applications
Information security policy includes detailed provisions for AI
applications, though it may not cover all potential scenarios or
emerging threats
The security policy acknowledges
AI but lacks comprehensive
provisions tailored to its specific
risks and operational needs
Governance &
guardrails
Model training &
finetuning
Cyber & privacy
transparency
● Curating AI privacy guidelines as per
regulations (E.g.,GDPR, AI Act)
● Adoption or custom design of AI cyber
risk management frameworks (E.g.,
Google SAIF)
● Sequentially train filtering policies
using supervised finetuning, reward
modeling techniques
● Solutions are trained on curated or
licensed content to detect IP misuse,
plagiarism, etc.
● Conduct regular 3rd party assessments
for the organization with specific
considerations around GenAI, data
privacy, etc.
● Input sanitization & prompt rate
limiting to be incorporated
GDPR: General Data Protection Regulation Google SAIF: Google's Secure AI Framework
~30% service players have a comprehensive security policy for AI; Exemplars
are focusing on building the right guardrails, finetuning models & transparency

56
Data | Overview
0 10 20 30 40 50 60 70 80 90 100
Average score
(out of 100)
Data Inventory, Storage
and Accessibility
A
Data Quality
Management and
Cataloging
B
Data Labeling, Metadata
and Annotating
C
Clearly defined data
ownership &
stewardship roles
D
E2E encryption to
secure data
E
65
64
69
72
75
xx

57
Data
Players have a good foundational base of readiness in terms of data storage
& classification, DQM & cataloging, and data labeling
Data Inventory, Storage and Accessibility
56%
18%
4%
22%
Comprehensive
centralized storage
of data sources
Formally classified
centralized inventory
of data sources
Centralized storage
of data sources
under development
Lack of centralized
storage of
data sources
Since we have large & rich repository of
consistent in-house data over the last
9-10 years, we are able to train the AI
models better
GCC player
Data Quality Management and Cataloging
20%
54%
26%
Advanced automated
systems deployed
for data quality
assessment
Established systems
for data quality
assessment
Basic mechanisms
for data quality
assessments
We often observe client data is
inconsistent & in varied formats. Data
standardization & cleanup is a potential
revenue generating offering for us
Large Tech player
Data Labeling, Metadata and Annotating
34%
42%
22%
2%
Well integrated,
sophisticated systems
Effective systems for
data labeling, metadata
& annotation
Some non –
comprehensive
systems being trialed
Absence of data
labeling, metadata &
annotation processes
We helped a client categorize and label
data accurately by tagging multiple
transactions by a single user to the
same user's profile instead of creating
multiple profiles for each transaction
BPO player

58
GenAI automates process of assigning tags/categories to unstructured data, such as images, text, or audio for various
applications like content filtering, sentiment analysis, & object recognition
Data labeling
& classification
Data cleansing with GenAI involves the automated process of identifying and correcting or removing errors,
inconsistencies, and inaccuracies in datasets
Data cleansing
automation
Automatic monitoring of model & output drift utilizing synthetically generated content & results to cross-check model
output & performance
Data quality
automation
Auto MDM (Master Data Management) powered by GenAI intelligently identifies and reconciles data, reducing
manual effort and improving data quality
MDM
automation
It involves creating artificial data that mimics real-world datasets. The benefits include preserving data privacy by
replacing sensitive information, enabling safe and legal data sharing
Synthetic data
generation
Allows users to analyze and visualize data independently, to gain insights and make informed decisions without the
need for extensive technical expertise or assistance
Augmented
analytics
Involves transforming personal or sensitive information in a dataset to ensure confidentiality and privacy while
preserving the utility and integrity of the dataset, allowing secure analysis and sharing of information while
protecting individual privacy
Data
anonymization
Case Study | Players starting to leverage GenAI to accelerate their Data
Transformation journey across multiple use cases
Use case Description
Data

59
Players have E2E encryption for most AI data;
Exemplars have MSAs, in-house encryption tools, etc.
MSA: Master Service Agreement, SoW: Statement of Work
End-to-end
encryption
is rigorously
enforced for all
AI data
End-to-end
encryption
enforced for
most AI data
End-to-end
encryption is
enforced in
some areas
In order to enhance compliance with E2E encryption practices,
Exemplars have adopted several measures:
E2E encryption
44%
38%
18%
Defining encryption policies, incorporated into contracts if required.
E.g., a Mid Tech player is re-writing client MSAs to include data
security clauses relevant for AI projects
Deploying techniques like SMPC and federated learning to
allow for model computations on encrypted data
Performing third party audits and validations of
encryption measures
Building in-house encryption tools for clients, including automation
of encryption (if required) E.g., a Large Tech player has developed
an in-house data encryption tool to mask data. It even allows
differential access rights across the client organization (employee
personal data only viewable to HR)
Data

60
Data ownership roles being defined clearly to ensure compliance with data
ownership norms; Exemplars enforce via BU level Chief Security Officers
Establishing clear data ownership norms helps define the rights, responsibilities, and expectations regarding the data used and generated
during the AI project
Key learnings from industry Exemplars:
Governance through
BU level CSOs (Chief
Security Officers)
Define risk mitigation
measures early-on to
ensure swift response
to data breaches
Specify data usage
norms that explicitly
state how the data will
be used within the
scope of the AI project
IP norms to address
ownership of AI
models, algorithms,
and other IP
developed during
the project
Data storage norms
outline where the data
will be stored, who has
access to it, and under
what conditions
Data retention and
deletion norms ensure
that data is not retained
longer than necessary
& compliance with
privacy regulations
Typical concerns on AI linked
data ownership
Defined data ownership & stewardship roles
● Storage & usage of client
specific training data
(post utilization for AI
model training)
● Data usage for co-pilots
during inferencing
● Indemnity clauses &
liabilities for any
data breach
● Multiple geography based
regulatory compliance
Comprehensive
framework for data
ownership
Well-communicated
roles for effective
data governance
40%
Data ownership
clearly identified for
most datasets
Clear understanding
of responsibilities
36%
Initial
efforts to
define data
ownership
24%
Clear data ownership
& definition of access
rights eliminates
data misuses not
just in Tech partners'
ecosystem but also
within client's
own ecosystem
Large Tech player
Data

61
Articulating the game plan: Key learnings from Exemplars
Building a winning client ecosystem
Developing AI offerings for clients
● AI offerings aimed at what clients are
buying, i.e. willing to pay for and offerings
that can generate the highest value at scale
● Light touch co-pilots with clients on areas
of interest for experimentation
Organization wide commitment to AI
● Dedicated oversight with cross-
functional stakeholder involvement in
value assessment
Clarity of vision for workforce
● Identify skillset required to accelerate
AI maturity
● Manage discomfort & uncertainties within
workforce through active resistance
management
Targeted GTM strategies
● GTM strategy across prioritized verticals,
accounts & use cases based on in-depth
understanding of client needs
Robust prioritization mechanisms
● Holistic frameworks across ROI,
feasibility & replicability to enable robust
prioritization of use cases including for
financing & production
CoE & ways of working
● Dynamic & evolving Center of Excellence
driving the AI agenda across the
organization & BUs
Attracting stellar talent
● Hiring the right talent across organization
levels with technical expertise &
business acumen
Agility in execution
● Improved TAT from identification
to production of use cases, rigorous
monitoring & relevant course correction
Leverage external partnerships
● Identify gaps in capabilities, develop
effective partnerships to expedite
AI maturity
Ethical AI policies
● Going beyond bias detection
incorporating privacy, security, etc.
Build a future ready workforce
Training curriculums aimed to
● Upskill engineers, data scientists, etc.
into AI experts
● Equip sales, pre-sales (other internal)
functions with AI expertise
Fostering AI awareness
● Engage early with clients through training
workshops, Art of the Possible demos, etc.
Building an organization for the future Grow by growing others

62
AUTHORS
Managing Director and Senior Partner
BCG
RAJIV GUPTA
Managing Director and Partner
BCG
SUDHANSHU CHAWLA
Senior Vice President and Chief Strategy Officer
NASSCOM
SANGEETA GUPTA
Partner
BCG
SAMBHAV JAIN
Senior Director and Head of Research
NASSCOM
ACHYUTA GHOSH
Director, NASSCOM Insights
NASSCOM
NAMITA JAIN
Project Leader
BCG
TRISHLA SELARKA
If you would like to discuss the themes and content of this report, please contact:

63
ACKNOWLEDGEMENTS
The authors thank and acknowledge the support provided by Divya Singhvi (Consultant), Khushi Kedia (Senior Associate) and Mayank Kak (Senior
Knowledge Analyst) in preparing this report.
We would like to extend our gratitude to NASSCOM member organizations, industry stalwarts and leaders from the IT & BPM sector for sharing their rich
experiences with us and enabling others to learn from their knowledge. Their expertise has been invaluable to this exercise.
A special thanks to India Marketing Team for managing the marketing process and to Saroj Singh, Sujatha Moraes, Vijay Kathiresan, Seshachalam Marella,
Soumya Garg, Aliviya Saha, Ratna Soni, and Saanchi Chatwal for their contribution towards design and production of this report.

64

65

66

ai-powered-tech-services-a-roadmap-for-future-ready-firms.pdf

More Related Content

What's hot

Similar to ai-powered-tech-services-a-roadmap-for-future-ready-firms.pdf

Recently uploaded

ai-powered-tech-services-a-roadmap-for-future-ready-firms.pdf