A 2023 Snapshot

June 2023
Leader’s Note
Generative AI, a subset of artificial intelligence (AI) that employs deep learning and neural network techniques on massive-scale data
to “create” novel content of varied types – text, images, video, audio, code, etc. – has taken the world by storm, and that too, at an
unparalleled pace in the last six months!

Globally, the phenomena got spotlighted with OpenAI’s $1 Bn equity infusion by Microsoft in 2019, but the real inflection point arrived
when GPT3.5 model was made public in November 2022. B2C applications have sprung up faster. While majority of the foundational
large language models, GANs, etc. are being built by the US-based startups and enterprises, there has been an exponential increase in
activity across the global startups since 2022, primarily in applications and services built atop these pre-trained models.
Sangeeta Gupta
Indian generative AI startup ecosystem has expanded rapidly since 2021, more than doubling in count, with over $475 Mn funding in Senior Vice President
the last 2 years. Majority of these Indian startups offer generative AI-native solutions, built in-house. However, a majority of them lack & Chief of Strategy
adequate funding. Unlocking the full potential of generative AI for India, and from India for the world, will not only require rigorous nasscom
assessment of sustainable whitespace opportunities in various sectors, but will also require innovative thinking towards inviting patient
capital and facilitating compute capabilities at scale.

Being a nascent technology, big-ticket investments, particularly in native Indian foundational models and enterprise-grade
applications services, are yet to happen. In addition to limited investments, generative AI startups in India also face the challenge of
limited high-quality and ready-to-use training datasets and lack of high-performance compute capacity at scale. Lack of clarity on
data privacy, security, ethical guidelines, and globally-consistent generative AI usage standards can further slowdown growth.

In preparing this first-look into the Generative AI Startup Landscape in India: A 2023 Snapshot, we surveyed and interviewed 50
startups, and researched nearly all active ones. The study highlights the overall state of growth and funding, core offerings, go-to-
market readiness, and key demand sectors for these startups. The study presents perspectives on major growth challenges and priority
recommendations to unlock advancements in this technology space. Both native and pivoted generative AI startups have been
included in the study.

Generative AI is potentially more transformative than all previous waves of AI, and it will take an ecosystem-level and a mission-scale
effort to build India’s niche in the global Generative AI landscape. We hope you find this first look into India’s generative AI startup
landscape useful!
Contents

04 07 12
Executive Summary Understanding Generative AI Global Generative AI Startup Landscape

17 28 32
Indian Generative AI Startup Landscape Indian Generative AI Startups’ Growth Outlook Appendix
Executive Summary – Global Generative AI Startup Trends and Drivers
Global Generative AI Trends Key Drivers of Generative AI Market Growth

Private investments in generative AI GPUs and TPUs Investment Inflection Natural Language
$2.5+ Bn startups in 2022 in 2010s Post 2019 Interface

Of global AI private investments of § Deep neural networks § Microsoft’s $1 Bn § Unlike artificial narrow
3% ~$92 Bn in 2022 invented in the 1980s, but investment in OpenAI in intelligence solutions,
limited by availability of 2019 provided the much- generative AI models and
Private generative AI investments compute resources needed impetus to applications democratize
$11 Bn+ announced in Q1 CY2023* commercialize generative usage based on natural
§ Nvidia’s 2010 launch of
pre-trained transformer or language prompts,
the world’s first GeForce
GPT models enabling a highly-
Number of active generative AI 256 general processing
550+ startups worldwide
interactive everyday AI
unit (GPU) accelerated § Investments grew 3X
experience
cloud adoption and AI between 2019-2022
Global generative AI unicorns as of advancement
14 Q2 CY2023

Features of Generative AI

Salient Features Major Content Categories Key Challenges

Data-Centric Self-Learning Text Images 1. Compute-Intensive

Code
2. Not Explainable
Augmentation-Capable Prompt-Led Videos Audio Multi-modal 3. Data Privacy/ Security

*$10 Bn is Microsoft investment in OpenAI.

Source: Stanford HAI AI Index 2023, Dealroom, CB Insights, Forbes, nasscom analysis
Executive Summary – Indian Generative AI Startup Trends

$475 Mn+ Indian generative AI startup Generative AI startups in India

Zero
funding during 2021-23* building large foundational models

60+ Active generative AI startups as Startups that are not funded 70%
of Q2 CY2023
Growth in number of Gen AI Indian generative AI startups
1.5X Investments Limitations that are not market-ready ~80%
Indian startups during 2021-23*

Increase in funding in 2022, y-o-y, Native datasets and India-based

12X making 2022 the max-funded year AI hardware OEMs
Nil

52% Indian generative AI startups Native Indian language-based

Vernacular
focused on text, images, and video models

70%+ Of the startups in India are Gen AI Strategy and Potential Akin to DPI*1, India’s native Gen AI as a
native; make in-house solutions Product Focus Ecosystem Enablers AI models Public Good
Top Tech Services, Media, Entertainment,
Knowledge Services
First nation to have built draft Responsible
Demand Responsible AI guidelines AI
Sectors

*2023 data is until 29 May 2023. *1 DPI is digital public infrastructure

Source: Tracxn, 2023 nasscom Gen AI start-up survey
Executive Summary – Challenges and Growth Recommendations

Major Challenges Faced by Global/ Key Recommendations for Growth

Indian Generative AI Startups of Indian Generative AI Startups

Lack of high-quality, large-scale, localized Startups Investors

Data training data

§ Co-innovate with industry and cloud § Invest in concepts fulfilling major

partners to build quick scale whitespaces in the Indian context,
Scarce patient capital for cost-intensive
Funding foundational model work
but with global market potential
§ Conduct thorough impact
assessments to avoid legal impact § Scrutinize for responsible AI
implementation
High cost of compute resources, and no scaled § Build specialist talent pipeline
Compute domestic hardware OEMs or hyperscalers in through college partnerships § Invest with “patient capital”
pipeline
Industry Governments
Data privacy, security, copyright infringement
Regulations concerns and no consensus on global regulations
and usage standards § Create awareness drives to § Identify high-priority use-cases
demystify generative AI on lines of DPI-scale adoption
Limited supply of new gen AI skills, difficult to § Champion and evangelize § Mission-scale funding for India-
Talent upskill a large workforce rapidly in the next 6-12 responsible AI framework based compute and native
months datasets
§ Drive co-innovation and open
innovation models § Enact the data privacy law
Disrupted net-zero goals, and the need to
Impact rebalance growth with current and emerging ESG § Partner with industry associations to § International consensus on
compliances upskill incoming talent at scale globally-consistent responsible AI
norms
Source: nasscom analysis
Understanding Generative AI #1
8 Generative AI Startup Landscape in India: A 2023 Snapshot

Generative AI Leaders’ Speak

Task-Centric to Data-Centric AI: Early Adopters Will Win Big:

“We’re seeing a shift around the world from task-centric AI to data-centric “It's not a crowded market yet, but it is getting more competitive. The
AI that supports multi-tasking and innovative capability.” barrier for entry is still quite low.”
- Co-Founder and CEO of a Gen AI startup - Founder and Managing Partner of a VC firm

Security, Privacy and Customization Still Nascent:

Supervised Classification to Free-Flowing Interaction:
“As enterprises define their generative AI strategies, they’re looking for
“The last decade has been defined by classification and prediction, now
privacy, transparency, customization and ease of deployment. Current
we’re looking at interaction. Rigidity of format is going to fall away and
cloud offerings, with closed-source models and data, do not meet their
everything is going to feel more dynamic and personalized.” requirements.”
- CEO and Co-Founder of a Gen AI startup - Co-Founder & CEO of a Gen AI startup

A Co-Pilot Unleashing Superior Productivity:

Data is the Real Moat:
For me, where generative AI can really shine is unblocking you, getting new
ideas and just getting you an assistant that is willing to do whatever you “On a macro level, I don’t think the moat is in the AI, at least not in the
want, 24/7.” models. I think there is a moat around the data.”
– Co-Founder of a Gen AI unicorn - CEO & Co-Founder of a Gen AI unicorn

Gen AI a Hype or Reality?:

Real Value from Data:
“The hype surrounding generative AI may be justified.”
“I think that in the future, most companies will be LLM-based. These tools
– General Partner at a VC firm
will enable them to generate real value from their own data.”
- CEO & Co-Founder of a Gen AI startup

Source: nasscom analysis

9 Generative AI Startup Landscape in India: A 2023 Snapshot

AI to Generative AI: Evolving Characteristics and Technology Stack

Key Characteristics of Generative AI High-Level Generative AI Tech Stack
Apps
Data-Centric Users
Models Apps
Infrastructure

Subset of AI that generates unrestricted, novel, varied types of § Applications and/or tools that integrate generative
Apps
content from existing data, beyond AI’s task-specific adoption End-user facing B2B and B2C applications without
AI models into user-facing systems, via either end-
thus far proprietary models to-end apps with in-built models or open APIs
Examples: Jasper, Github Copilot
§ Available for diverse categories across text, image,
Self-Learning video, code, or a combination of these
End-to-End
Employs deep learning, a subset of machine learning, in Apps
Model Hubs
unsupervised or semi-supervised constructs, self-corrective End-user facing
Platforms to share and
host models
Models
Closed-Source
training applications with
Foundation Examples: Hugging Face,
proprietary models
Models Replicate § Core algorithms trained using deep learning
Examples:
Augmentation-Capable Large-scale,
techniques on large datasets to learn patterns by
Midjourney,
Runway pre-trained discriminative or generative methods.
models exposed
Ability to employ backward and forward propagation in the Open-Source
via APIs Foundation Models § Variational Autoencoders (VAEs), Generative
same learning model to “augment” existing data in unique Examples: GPT3 Models released as trained Adversarial Networks (GANs), Transformers, etc. are
ways (OpenAI) weights majorly deployed generative AI model types
Examples: Stable Diffusion
Compute-Intensive (Stability) § Models can be closed or open-source

Requires high-performance, massively parallel processing tech

stack and specialized training models, API libraries and Cloud Platforms
Infrastructure
visualizations Compute hardware exposed to developers in a cloud deployment model
§ High-performance hardware and cloud platforms to
Examples: AWS, GCP, Azure, Coreweave
run intensive compute and store tasks in large-scale
Unexplained model training
Closer to general-purpose technology tag, generative AI is still Compute Hardware
§ Data pre-processing tools, model serving systems,
based on largely unexplained models, with a higher risk of Accelerator chips optimized for model training and inference workloads software frameworks and libraries are included to
inherent bias Examples: GPUs (Nvidia), TPUs (Google) facilitate the development, training, and deployment
of generative AI models

Source: Andreessen Horowitz, nasscom analysis

10 Generative AI Startup Landscape in India: A 2023 Snapshot

Generative AI: A Gradual Shift to Artificial General Intelligence

Artificial Narrow Intelligence (ANI) Generative AI Artificial General Intelligence (AGI)

Narrowly-trained algorithms focused on specific, Pre-trained models that perform a broad set of
Intelligent agents potentially capable of defining
repetitive tasks defined by human operators; self- pre-defined tasks, from generating text to images,
Purpose tasks, pathways to task completion, and en-route
expansion to solve unfamiliar problems is nearly videos, code, etc. based on simple human
self-correction, comparable to human cognition
non-existent language prompts

Models are supervised or unsupervised, fixed- Generative models, such as large language models AGI models could potentially be self-selected
Learning Model domain, pre-programmed, on largely structured (LLMs), are unsupervised or semi-supervised, use learning models based on time, data, and output
data unstructured or structured data context

Lacks self-awareness, consciousness, and creative Pre-trained self-awareness, inability to Self-aware, possesses common sense, creativity,
Awareness
cognitive abilities demonstrate common sense, creative and ability to expresses emotions

Uses labeled or annotated data and specific

Primarily uses deep learning techniques on Near-human, or beyond, generalization
algorithms based on computer vision, NLP,
Data Processing unlabeled data using backward propagation to capabilities across variety of inputs based on
machine learning, or neural networks for
learn data “augmentation” capabilities massively-parallel real-world data processing
classification/segmentation

Maturing global B2B/B2C usage to detect

Fastest B2C adoption to generate content, art, etc.
Adoption Stage patterns, make/support decisions, hone analytics, Primarily in R&D across a limited set of countries
Enterprise adoption nearly non-existent till date
classify data and detect fraud

Supports repetitive tasks and process automation, Empirical tech, potential for deep bias with block- AGI is at risk of pre-mature commercialization and
and predictive decision making along with box models, creative but inaccurate output, and long-term implications if quick wins subsume
Implications
advanced analytics, but has invited concerns on exponential energy consumption – has invited broad-based general intelligence evolution, and
bias and AI explainability sharp regulatory discussions unsustainable energy consumption

Source: nasscom analysis

11 Generative AI Startup Landscape in India: A 2023 Snapshot

Generative AI Evolution Timeline: Q12021 – Q12023

September 2022 March 2023
March 2023
NVIDIA releases Get3D, September 2022 Google announces
Jan 2021 OpenAI releases
a 3D-shape generator Meta announces generative tools
OpenAI launches Dall-E ChatGPT on GPT4.0
trained only on 2D Make-A-Video integration into Google
model
images Workplace

May 2021 July 2022 October 2022

Google announces BLOOM debuts on the Google launches February 2023
LaMDA HuggingFace Imagen Video Google releases BARD

June 2021 November 2022

July 2022 February 2023
GitHub announces its OpenAI launches
Stability.ai launches Google invests $300 Mn
Copilot tool ChatGPT on GPT3.5
Stable Diffusion in Anthropic
model

January 2023 February 2023

December 2021 May 2022 Microsoft releases beta
Microsoft announces
Baidu releases its ERNIE Meta releases Open upgrade to Bing search
$10 Bn investment in
3.0 Pretrained Transformer with summarization
OpenAI, exclusive Azure
(OPT) model and chat based on
integration
ChatGPT

April 2022
April 2022 February 2023 February 2023
Google’s releases
OpenAI releases Dall-E AWS and HuggingFace Runway.ai announces its
Pathways Language
2. announce partnership Gen-1 model
Model (PaLM)

Source: PitchBook
Global Generative AI #2
Startup Landscape
Key Global Generative AI Investment Trends in 2022-2023

~$92 Bn 14 116 32%

Total global private investment in AI Global Active Global startups
Generative AI institutional have Seed A or
in 2022
unicorns investors in 2022 higher stage
funding

10X+ $2.5 Bn+

550+ $48 Bn
Generative AI Valuation of top 100 Valuation of Total global investments in
startups worldwide global generative AI OpenAI, at $29 generative AI in 2022, 3% of
as of Q2 CY2023 startups Bn, over the next
biggest unicorn total AI private investments

Source: CB Insights, Dealroom, Forbes, Stanford HAI AI Index 2023, Tracxn, nasscom analysis
14 Generative AI Startup Landscape in India: A 2023 Snapshot

Generative AI Funding and Valuations Hit Inflection in 2022

Top 100 startups collectively valued at nearly $50 Bn

Cumulative Funding of Global Generative AI Startups Total Valuation of Major Global Generative AI Companies
2021-2023P (in $Bn) Illustrative logos represent unicorns

19.7*
50

40

30

20
7.4
4.9 10

0
2021 2022 2023P

Other Company Evaluation of at least $0.2 Bn (but not unicorn)

2021 2022 2023 P
OpenAI Company Evaluation of at least $1 Bn or more (unicorn)

§ Total private investments grew almost 4X between 2021 and 2023 (till § Combined enterprise value of 100 key generative AI companies stood
date) at $48 billion in January 2023, marking a 7x increase from 2019, when
the first set of generative AI unicorns emerged
§ 2023 could witness maximum investments till date in any year, with $10.68
Bn worth committed in Q1 itself § OpenAI alone accounts for 50% of the above total valuation, reaching
$29 Bn in its valuation in Q1 2023
§ Deal volume stabilized in 2022, though per deal funding grew by more
than 40% during 2021-22, suggesting more mature/growth stage funding § 14 generative AI startups reached unicorn status, around 50% in just
in proven generative AI startups 2022

*$10 Bn is Microsoft investment in OpenAI.

Source: CB Insights, Dealroom, nasscom analysis
15 Generative AI Startup Landscape in India: A 2023 Snapshot

Global Generative AI Investments in 2021-22 are Spread Across Diverse

Categories
65% flowing into text and visual media applications

Visual Media Graphic & Product

Design
Hyper-Real
Deepfake Videos
Generative Interfaces
$822M | 58 Deals $22M $15M
$586M | 20 Deals

Consumer Enterprise AI Images & NFTs General Human-Machine

Social Avatars $114M Search Interfaces
Apps &
Tools $253M $85M $290M
Photo & Video Synthetic
Lip Sync & Video
$133M Editing Tools Training Data
Personalization
$4M $131M $126M

Motion Capture
& Animation Productivity & Knowledge Management
$14M $211M
Cross-Functional
(APIs) Social Media & Summarization
Marketing Content
$199M Documentation
$249M Code
Audio Editing Music
$317M Generation Completion $3M
$80M $47M $103M
Text-to-Code

Voice Voice Synthesis $4M

Dubbing & Cloning
Consumer Support Comd. Line Web & App
$65M $42M $43M $19M
$10M
Text Personal Writing
& Storytelling
Sales, email & Outreach Speech & Audio Code
$852M | 48 Deals $6M
$15M $212M | 26 Deals $140M | 20 Deals

Source: CB Insights
16 Generative AI Startup Landscape in India: A 2023 Snapshot

An Illustrative Global Generative AI Landscape Indicates Rapid Ecosystem

Acceleration
Particularly at the hardware and cloud layers
Illustrative

Chips
1 Infrastructure
Cloud

Text LLMs/NLP Image Models Video Models Audio & Speech Tools & Services

2 Foundational
Models

General General Sales & Coding Customer User Knowledge General Consumer Design Games Music Full Stack
Search Writing Marketing Service Interface Manage/ Image & Avatar & Audio & Apps
B2B-B2C Chatbots /RPA Summary/ Vedio Face Apps Speech (Legal,
Search Tools Biotech,
3 Applications Finance,
etc.)

Design –
Fashion, Finance & Tech Publishing & Film, Music, Legal & BioTech/
4 Industry Products, Insurance & Telco Advertising Radio, Creative
Games
Compliance Pharma
Architecture

Source: CB Insights, BASE10 PARTNERS, nasscom analysis

Indian Generative AI Startup #3
Landscape
India’s AI Startup Ecosystem and Potential, 2022-23 Trends

$590+ Mn 60+ 79% ~80%

Total private investments in Indian generative Indian generative
Generative AI
AI startups are AI startups are not
generative AI Indian startups, till Startups in India,
building solutions market ready
date, $475 Mn since 2021 Q2 2023
in-house

~$8 Bn
70% 30% 70%
Funding came in Generative AI Are generative Total private investments into AI
2022 alone startups in India AI native (not from 2013-2022, $3.24 Bn in
have received pivoted) 2022 alone across 1900+ AI
funding startups in India

Source: Stanford HAI AI Index 2023, nasscom State of Data Science & AI Skills in India, 2023 nasscom Gen AI start-up survey
19 Generative AI Startup Landscape in India: A 2023 Snapshot

Indian Generative AI Ecosystem Landscape Shows Rapid Uptick in the

Applications and Services Space
However, with a glaring gap in native foundation models Illustrative

Chips
1 Infrastructure
Cloud

2 Foundational Text LLMs/NLP Image Models Video Models Audio & Speech Tools & Services
Models
No Indian Foundational Models Exists

General Multilingual Sales & Coding Customer User Knowledge General Consumer Design Games Music Full Stack
Search Dubbing Marketing Service/ Interface Manage/ Image & Avatar & (Fashion, Audio & Apps
Applications – General Chatbots / RPA Summary/ Video Tools Face Apps Architect) Speech (Legal,
Majority Indian Writing Search Biotech,
3 Generative AI B2B-B2C Finance,
etc.)
Startups

Generative AI
4
Services

Design –
Fashion, Publishing & Legal & BioTech/
5 Industry Products,
Manufacturing BFSI Tech & Telco
Advertising
Entertainment
Compliance Pharma
Games
Architecture

Source: nasscom analysis

20 Generative AI Startup Landscape in India: A 2023 Snapshot

Number of Startups in Indian Generative AI Landscape Doubled in 2022

Cumulative Number of Indian Generative AI Startups Illustrative logos represent

Gen AI companies present in the landscape
Cumulative funding till end of year/ till date

$598 Mn+
$566 Mn+ 60+
55+

Key Themes
$157 Mn+
§ Generative AI startups in India more than
25+ doubled between 2021 and 2023 (till date)
§ Indian generative AI start-ups had already
raised $590+ Mn in cumulative funding till
May 2023; 80% of this funding, $475+ Mn,
was received from 2021 onwards
§ Incumbent startups garnered over 90% of
the total investments
§ Diversification across categories has gained
momentum since 2022, with more startups
targeting data-driven whitespaces in
text+image, image+video, audio+video, etc.

2021 2022 2023 P

Source: Tracxn, nasscom analysis

21 Generative AI Startup Landscape in India: A 2023 Snapshot

At 45%, Bengaluru Leads All Major Locations in the Number of Generative

AI Startups, Followed by Mumbai and Delhi/NCR Regions

§ Greater Bengaluru region has the largest number of generative

AI startups in India, 45%. The city’s overall environment,
including deeptech, startup ecosystem, high-end innovation-
driven institutions, extensive industry presence, and an
emerging class of domestic angel investors, are a big draw
§ Mumbai and Pune regions offer the second largest pool at 21%. Delhi-NCR
This region boasts of some of the most well-established 10%
institutional investors and VCs, and a diverse talent pool
§ Delhi NCR benefits from strong presence of the knowledge-
based advisory industry and BPM enterprises. Startups with Ahmedabad
applications suited to these categories further benefit from the 5%
innovation ecosystem of established educational institutions
and a large pool of college graduates
§ Hyderabad offers a nation-leading innovation infrastructure, Mumbai/Pune Hyderabad
particularly for the deeptech ecosystem 22% 9%
§ Chennai offers advantages of being the country’s leading SaaS
hub, with strong presence of product organizations, as well as
focused innovation and R&D labs of the global capability
centers in India Bengaluru Chennai
45% ~9%

Source: 2023 nasscom Gen AI start-up survey

22 Generative AI Startup Landscape in India: A 2023 Snapshot

3-in-4 Startups in India are Generative AI Native Companies

With as many building their products in-house

Native vs. Pivoted

Generative AI Product Development Models
Generative AI Startups

Collaboration/ Project
Partnership Basis
10% 11%
Pivoted
26%

Gen AI
Native
74%

In-house
79%

§ 74% of startups are generative AI native. § 79% of the Indian Gen AI startups prefer to build their
§ Majority have established between 2021-22, reflecting the the solutions in-house.
confidence of companies in the growth potential of the § While this shows the risk appetite of these startups, with
technology. investors turning value-driven, it is imperative for startups to
find the right whitespaces.

Source: 2023 nasscom Gen AI start-up survey

23 Generative AI Startup Landscape in India: A 2023 Snapshot

78% of the Generative AI Startups in India are Yet to Commercialize

their
With as Products
many expecting to go to the market in 1-3 years

Generative AI Startups By Product Development Lifecycle Stage and Commercialization Timelines

Prototype/MVP
1-2 years
23%
18%
2-3 years
5%

Commercialized Within 1 year For Non-

Proof of Concept
22% 48% Commercialized
37%
Unsure
29%

Development
18%

§ Around 80% of Indian generative AI startups are yet to commercialize their offerings
§ 60% are at the proof-of-concept or prototyping stage
§ 37% of the non-commercialized solutions are likely to find markets within an year, as most applications layer startups face time-product-market
fit pressures

Source: 2023 nasscom Gen AI start-up survey

24 Generative AI Startup Landscape in India: A 2023 Snapshot

Indian Generative AI Startups Offer Solutions Applicable Across Major

Categories
Similar to the global peers, with applicability across sectors

Trends in Generative AI Solution Categories Use Case Top Industries

Text Content Creatio n 18% Text Content Creation Media, E-commerce, Advertising, Tech Services

Chatbots and Virtual Assistants 18% Chatbots and Virtual Assistants Banking, E-commerce, Tech Services

Im age and Video Generation 16% Image and Video Generation Media, Entertainment, Advertising

Predictive Modeling 14% Predictive Modeling Finance, Healthcare, Retail

Language Translation 11% Language Translation E-commerce, Travel and Hospitality, Tech Services

Fraud Detection Fraud Detection Banking, Insurance, E-commerce

8%

Autonomous Vehicles Automotive, Transportation, Logistics

Autonom ous Vehicl es 5%

Fashion and Product Design Retail, Fashion, E-commerce

Fashion and Product Design 4%

Game Developm ent Game Development Gaming, Entertainment

3%

Drug Discovery 3% Drug Discovery Pharmaceutical, Life Sciences, Biotechnology

Source: 2023 nasscom Gen AI start-up survey

25 Generative AI Startup Landscape in India: A 2023 Snapshot

Majority Indian Generative AI Startups Prefer a Tech Stack Comprising

Public cloud and cloud-based databases, pre-trained models, and custom-built visualization tools

Matplotlib
Majority startups are inclined towards custom visualization tools that offer the
Apps and Plotly flexibility to design interfaces for specific use cases or data types, without
Visualization
Tableau having to use integration across multiple OTS tools

GPT Startups that launched before 2022 use closed-source pre-trained models by
default. However, there is an upcoming set of startups that are building
Models BERT
integration platforms to provide a single-click access to multiple open and
GLaM
closed-source models through API-based toolkits

MongoDB Majority AI applications continue to be built on relational databases, which is

changing with generative AI that relies on non-relational or vector databases.
Databases PostgreSQL
Non-relational databases, such as MongoDB, are capable of fast parallel
MySQL
processing across unstructured data

AWS
Public cloud is the preferred choice for most generative AI startups, unless they
Infrastructure Azure are training on proprietary data with API-based connect to private data sources
GCP or in-house data centers

Source: 2023 nasscom Gen AI start-up survey

26 Generative AI Startup Landscape in India: A 2023 Snapshot

Despite Rapid Acceleration in the Number of Generative AI Startups

Only 30% are funded, although 80% of these startups expect investments in 2023

Funding Raised in 2022 Funding Expectations in 2023

$5 Mn - $20 Mn $500k - $1 Mn
3% 24%

$500k - $1 Mn
7%

$1 Mn - $5 Mn No funding
7% raised yet $1 Mn - $5 Mn No plan
70% 43% 19%

Less than $500k

13%

$0 - $500k
14%

§ 70% of Indian generative AI startups continue to be unfunded, 2X+ that of global 33% that are without funding
§ Top funded companies include Observe.AI, Instoried, Magnifi, MadStreet Din, and Pepper Content
§ 80% of the startups expect to raise some funding in 2023, nearly half expecting it in the $1-5 Mn range

Source: 2023 nasscom Gen AI start-up survey

27 Generative AI Startup Landscape in India: A 2023 Snapshot

Growing Interest of Institutional Investors in Indian Generative AI

Startups
More than 50% of these institutional investors are India-based

Institutional vs Corporate Investors Domestic vs Foreign Investors

Corporate
Investor
28%
Foreign
43%

Domestic
57%
Institutional
Investor
72%

§ Active Corporate Investors: Rosebay Consulting, Cosme Matias § Active Domestic Investors: Venture Catalysts, Titan Capital,
Menezes, Zoom, Fractal, Times Internet, etc.* 9Unicorns, Mumbai Angels, Kalaari Capital, etc.*
§ Active Institutional Investors: SOSV, Venture Catalysts, Titan § Active Foreign Investors: SOSV, Artesian, Y Combinator,
Capital, Artesian, Nexus Venture Partners, etc.* BPEA, Alpha Wave Global, etc.*

Source: Tracxn
* Illustrative
Indian Generative AI #4
Startups’ Growth Outlook
29 Generative AI Startup Landscape in India: A 2023 Snapshot

India has Punched Below its Weight in the Overall AI Market

And the emerging generative AI arena, compared to the USA and Israel, two of the top-five global technology startup hubs

India USA Israel

Total Funding in AI, 2013-2022 $8 Bn+ $249 Bn+ ~$11 Bn+

# of Generative AI Startups, May 2023 60+ 400+ 70+

Total Funding in Generative AI Startups, till date $590 Mn+ $22 Bn+ $1 Bn+

Generative AI Unicorns, till date Nil 14 Nil

Foundation Models, till date Nil 30+ 2+

# of generative AI startups funded, disclosed 30%+ 65%+ 60%+

Source: Start-up Nation Central, Tracxn, Stanford HAI AI Index 2023, nasscom analysis
30 Generative AI Startup Landscape in India: A 2023 Snapshot

In Addition to Globally Applicable Concerns with Scaling-up Generative

AI Solutions
Indian startups face major limitations with data, funding, and compute resource availability

Data Lack of high-quality, large-scale, localized training data

Funding Scarce patient capital for cost-intensive foundational model work

Key
Challenges
Compute High cost of compute resources, and no scaled domestic hardware OEMs or hyperscalers in pipeline
in Scaling
Generative
AI Startup Data privacy, security, copyright infringement concerns and no consensus on global regulations and usage
Regulations
Ecosystem standards
in India

Talent Limited supply of new gen AI skills, difficult to upskill a large workforce rapidly in the next 6-12 months

Impact Disrupted net-zero goals, and the need to rebalance growth with current and emerging ESG compliances

Source: nasscom analysis

31 Generative AI Startup Landscape in India: A 2023 Snapshot

Recommendations and Resulting Execution to Grow the Indian

Generative AI Startup Ecosystem Will Have to Move with the Speed of
Global Developments in this space
Startups Investors

§ Co-innovate with industry and cloud partners to build quick scale § Invest in concepts fulfilling major whitespaces in the Indian context,
but with global market potential
§ Conduct thorough impact assessments to mitigate potential legal
impact § Scrutinize for responsible AI implementation
§ Build specialist talent pipeline through academic partnerships and § Invest with “patient capital”
internship opportunities

Industry Governments

§ Create awareness drives to demystify generative AI for consumers § Identify high-priority use-cases across government departments,
and businesses focusing on integration with public services and existing DPI
platforms
§ Champion and evangelize responsible AI framework
§ Mission-scale funding for local compute resources, Indian datasets,
§ Drive co-innovation and open innovation models in the space of
and innovation in the space of quantum and AI
applied generative AI
§ Enact the data privacy law
§ Partner with industry associations, such as nasscom and nasscom’s
FutureSkills Prime platform to skill incoming talent at scale § Lobbying in international fora for globally-consistent responsible AI
guidelines

Source: nasscom analysis

Appendix #5
33 Generative AI Startup Landscape in India: A 2023 Snapshot

Sample List of Indian Generative AI Startups (1/6)

Company Product Category Website Link Description

Customer
Aspiro https://letsaspiro.com/ Sales adaptation tool providing instant feedback and scores for reps based on buyer personas
Service/Contact
AI-powered image editing tool that enhances product photos with ease to create professional-looking sales
Blend Image https://www.blendnow.com/
content like Social Graphics, Product Photos and professional SEO-optimised copies

Gen AI tool that utilizes cutting-edge technologies, such as NeRF for video to 3D, GAN for Text to 3D, and
brahmGAN Image https://www.brahmgan.ai/
Blockchain to democrotize 3D content creation

Dubverse Audio https://dubverse.ai/ Dubverse makes content multilingual with a single click to reach a broader audience.

Epowerx Multimodal https://epowerx.ai/ Epowerx extracts data from invoices, along with style transfer in images and personalized videos.

Hexo.ai Image https://www.hexo.ai/ Hexo allows generation of high-quality lifestyle photos for products in seconds

NeuralGarage Video https://visualdub.ai/ The solution synchronizes lip movements with audio using Gen AI technology

NeuroPixel.ai Image https://www.neuropixel.ai/ The tool streamlines and simplifies fashion catalogue creation using image generation technology

Portkey makes it easier to launch production-ready applications using LMOps stack for monitoring, model
Portkey.ai Tool/Models https://portkey.ai/
management, and more.
ViXR offers end-to-end solutions for creating, distributing, presenting, and analyzing 3D/XR product
ViXR.io Multimodal https://www.vixr.io/
demonstrations
Customer
Vodex.ai https://vodex.ai/ Vodex.ai provides an easy-to-use software for automating outbound phone calls.
Service/Contact
Zapero leverages generative AI to analyze market trends and generate clothing designs with simple text
Zapero.ai Image https://www.zapero.ai/
prompts
Customer DaveAI offers digital sales avatars that answer customer queries, recommend the right products, and offer
DaveAi https://www.iamdave.ai/
Service/Contact personalized experience to customers

Source: nasscom analysis

34 Generative AI Startup Landscape in India: A 2023 Snapshot

Sample List of Indian Generative AI Startups (2/6)

Company Product Category Website Link Description

Velocity Chatbot https://www.velocity.in/ Velocity’s Lexi is a ChatGPT-based chatbot that offers business insights to ecommerce companies

Spyne.ai Image https://www.spyne.ai/ Spyne AI enhances online car selling experience with AI-powered photography and editing.

Haptik offers an AI-powered chatbot that provides features such as situation-based answers, bot variant
Haptik Chatbot https://www.haptik.ai/
generator, natural document answering, and intent enhancer

Hype AI creates e-commerce stores based on best practices and design trends, offering personalized
Hype Software https://www.hype.store/
recommendations for a better shopping experience.

SigTuple Data/Reports https://sigtuple.com/ Sigtuple leverages generative AI to generate lab reports from the analysis of samples tested.

Fluid AI Chatbot https://www.fluid.ai/ Fluid AI's GPT is a tailored GPT for enterprises that understands an organizations content

The Artiste is a sophisticated deep learning-based audio-visual platform for generating high-quality 3D
Kroop AI Multimodal https://kroop.ai/ videos with digital avatars using audio or text inputs. Kroop’s VizMantiz API allows for detection of
unethically generated synthetic media.

Maya Labs Code https://mayalabs.io/ The tool uses Gen AI to assemble and deploy ready-to-use software based on text inputs in English.

https://www.neuralsyncai.com
Neural Sync Video Wav2Lip is a cutting-edge video AI tool for lip-syncing translated videos in multiple languages
/landing

PixAnimate is a ChatGPT-integrated tool that allows users to create videos in bulk and add natural language
PixAnimate Video https://www.pixanimate.com/
voice-overs to them

Creative AI Studio analyses social media and performance data to generate text and image-based
Pixis Multimodal https://pixis.ai/
creatives for various channels. The tool also allows text to video generation.

The company offers Gen AI tools and models for users to use and build their own products and services. The
A3logics Services https://www.a3logics.com/
team of experts at the company help users in developing their Gen AI models.

Source: nasscom analysis

35 Generative AI Startup Landscape in India: A 2023 Snapshot

Sample List of Indian Generative AI Startups (3/6)

Company Product Category Website Link Description

Pepper Pepper Content Marketing Platform enables fast content ideation, creation, and distribution using
Data/Reports https://www.peppercontent.io/
Content generative AI.

Mad Street The company offers an AI-powered solution for creating high-quality on-model product photos without
Image https://www.madstreetden.com/
Den elaborate photoshoots.

The company offers an end-to-end AI-based content writing tool that can easily generate and optimize
Instoried Multimodal https://instoried.com/
high-quality content

Contlo Image https://contlo.com/ Contlo offers an AI Marketing Assistant that is brand-specific not generic

Magnifi is an AI-driven platform for extracting highlights and key moments from video content for social-
Magnifi Video https://magnifi.ai/
ready clips in real time

vidyo.ai Video https://vidyo.ai/ The tool instantly creates short videos from longer ones seamlessly

Zocket Image https://zocket.com/ Zocket is an AI-managed ad creation and management platform for running ads on autopilot.

Scalenut Data/Reports https://www.scalenut.com/ Scalenut is an AI-powered assistant supporting the entire SEO content lifecycle.

Play.ht Audio https://play.ht/ The tool leverages Gen AI to generate realistic speech from input text.

LongShot Text https://www.longshot.ai/ The tool converts ideas into fully formed blogs within minutes.

Baselit offers a wide range of capabilities ranging from writing complex SQL queries and analyzing data
Baselit Data/Reports https://baselit.ai/
to creating dashboards and answering ad-hoc questions related to the data.

Listnr Audio https://www.listnr.tech/ The tool leverages Gen AI to generate realistic speech from input text.

Basch Video https://basch.io/ The tool creates captivating videos from written content without technical skills

Source: nasscom analysis

36 Generative AI Startup Landscape in India: A 2023 Snapshot

Sample List of Indian Generative AI Startups (4/6)

Company Product Category Website Link Description

Company’s Tradepilot.tech allows users to easily create fully automated algorithmic trading strategies and
Sciartis Labs Multimodal https://sciartis.com/
its GenAIPlayground allow users to experiment and create various media types using Gen AI.

The Gen AI-backed app allows users to manage entire family’s insurance policies. The app offers data-
InsurStaq Data/Reports https://www.insurstaq.com/
backed advice, insights from existing policies, and recommendations for new policies.

LeewayHertz Services https://www.leewayhertz.com/ The company offers Gen AI development services, including building of complete solutions, to its clients.

Cointab Data/Reports https://www.cointab.in/ The solution offers easy reconciliation of data and helps users in taking data-driven decisions.

CyberXGen Services https://cyberxgen.com/ The company offers cybersecurity technology, consulting & training services to its customers.

Prima Vision Data/Reports https://primavision.io/ The Gen AI-backed tool offered by the company is capable of creating structured reports from data

The AI engine uses Gen AI to turn millions of consumer conversations and reviews from across the web into
Bewgle Data/Reports https://bewgle.com/
deep insights.

The solution leverages its capability to analyze data to derive actionable insights from it which can help in
Crayon Data Data/Reports https://www.crayondata.com/
improving revenue, customer engagement, sales conversion rate etc.

The company’s products allow users to build multiple AI models in real time and deploy them as APIs
eizen.ai Tools/Models https://www.eizen.ai/ without writing any code. The solutions also allow users to monitor AI models and retrain them, and easy
interaction using eizenGPT.

Beatoven.ai Audio https://www.beatoven.ai/ The Gen AI solution allows users to generate customizable royalty free music.

Staqu Data/Reports https://www.staqu.com/ Jarvis derives insights from real-time video footages and presents them in a structured format to users.

Source: nasscom analysis

37 Generative AI Startup Landscape in India: A 2023 Snapshot

Sample List of Indian Generative AI Startups (5/6)

Company Product Category Website Link Description

TrueReach’s workflow agent, analytics agent, content agent, all-in-one assistant reimagine content
TrueReach.ai Multimodal https://www.truereach.ai/
creation, marketing, and customer engagement for brands.

United We
Chatbot https://www.unitedwecare.com/ United We Care’s Stella is a Gen AI-powered wellness coach fluent in 29 languages
Care

Wokelo AI Data/Reports http://www.wokelo.ai/ Wokelo is a Gen AI tool that can generate comprehensive research and analysis reports.

Verloop.io’s Generative AI platform allows businesses to transform their customer service function by
Verloop.io Text https://verloop.io/ offering them features such as conversation summary, message rephrasing, tone adjustment, and
expanding phrases.

Zpod.ai Text https://www.zpod.ai/ Zpod is a Gen AI tool that generates post ideas and post drafts to create effective LinkedIn posts.

Observe.AI generates insights from customer interactions, and enables teams to take smart actions and
Observe.AI Data/Reports https://www.observe.ai/
automate workflows

Rephrase.ai Video https://www.rephrase.ai/ Rephrase Studio is a text to video generator that builds digital avatars to make videos more professional

CodeMate Code https://www.codemate.ai/ The Gen AI-powered solution speeds up coding with the fastest error-free code writing capability.

Lease
Data/Reports https://leaseabstractai.com/ The tool automates lease abstraction and generates accurate reports in minutes.
Abstract AI

Postify quickly generates engaging posts for LinkedIn, Facebook, and Instagram based on images and
PostifyAI Text https://www.postify.ai/
some description

Source: nasscom analysis

38 Generative AI Startup Landscape in India: A 2023 Snapshot

Sample List of Indian Generative AI Startups (6/6)

Company Product Category Website Link Description

Patenti Data/Reports https://www.patentitech.com/ The Gen AI-backed tool is meant to ease the entire process of IP portfolio management

https://www.tarisa.tk/resolution
Tarisa Data/Reports The tool allows financial institutions to easily build credit profile of borrowers using Gen AI.
/tarisaweb.html

The company offers AI inference performance for frameworks including Tensorflow, PyTorch and ONNX
Ampere AI Tools/Models https://amperecomputing.com/
Runtime. The company also offers pretrained models to aid in accelerating AI development.
AI SEO
Text https://aiseooutlines.com/ The tool assists in writing SEO-friendly articles by providing all necessary information based on a keyword.
Outlines
https://www.vasanatechnologie The solutions offer multiple capabilities such as easy game development, and seamless translation of
Vasana Multimodal
s.com/ languages among others.
The company’s data exploration and report automation tools leverage Gen AI to enable enterprises in
vPhrase Data/Reports https://www.vphrase.com/
building a data-driven decision making.

Suki’s Gen AI voice assistant for healthcare offers features such as notes generation, taking dictation
Suki AI Text https://www.suki.ai/
commands, simplifying coding, answering questions, and deep EHR integrations.

Source: nasscom analysis

39 Generative AI Startup Landscape in India: A 2023 Snapshot

Major Generative AI Application Categories

Category Description

Image Computer-generated visual representations, including realistic images, abstract art, or graphic designs.

Text AI-generated written content such as articles, stories, poems, or news summaries using natural language processing.

Audio AI-generated audio content, encompassing music, speech, voiceovers, or sound effects for multimedia applications.

Video AI-created video content, spanning from short films, animations, or advertisements to visual storytelling and special effects.

Multimodal Generative AI products that combine two or more modalities such as generating outputs with text, image, audio, or video components.

Code Generated programming code or scripts automating tasks, solving mathematical problems, or implementing algorithms.

Tools/Models AI-developed software tools or models designed for specific tasks like image recognition, language translation, or recommendation systems.

Gen AI solutions that develop fully-functional websites or software based on user-provided information/requirements, including design and layout
Software
customization.

Generative AI technology services provided by companies to assist customers in building and deploying Gen AI software solutions with ease, akin to traditional
Services
technology service companies.

Customer Generative AI products offering outbound and inbound call capabilities, along with avatars that interact with users in natural language to provide solutions or
Service/Contact training.

Chatbot AI-powered conversational agents capable of simulating human-like conversations and providing automated customer support or information retrieval.

Data/Reports Generated datasets or reports offering comprehensive insights, analytics, visualizations, or summaries derived from various data sources.

Source: nasscom analysis

40 Generative AI Startup Landscape in India: A 2023 Snapshot

Key Generative AI Technology Stack Components

Category Description
Machine Learning Machine learning frameworks are comprehensive software libraries that provide tools and APIs for developing and training machine learning models. They
Frameworks offer functionalities for tasks like data preprocessing, model building, and optimization.
Programming Generative AI programming languages are programming languages commonly used for implementing and executing generative AI algorithms. They provide
Languages libraries, frameworks, and syntax that enable developers to create and manipulate generative models.
Web Scraping Web scraping tools and technologies are used to automatically extract data from websites. They enable developers to retrieve structured information by
Tools/Technologies interacting with web pages, parsing HTML code, and extracting relevant data for further processing or analysis.
Database Database tools and technologies provide efficient storage, management, and retrieval of structured data. They offer functionalities for creating, querying, and
Tools/Technologies modifying databases, enabling generative AI applications to store and access data effectively.
Data Processing Data processing tools and technologies facilitate the manipulation and transformation of data for generative AI workflows. They provide functionalities for
Tools/Technologies tasks like data cleaning, aggregation, filtering, and transformation, enabling efficient preparation of data for generative models.
Visualization Visualization tools and technologies enable the creation of visual representations of data. They provide capabilities for generating charts, graphs, and other
Tools/Technologies visual elements to help understand and communicate patterns, trends, and insights derived from generative AI processes.
Generative models are machine learning models designed to generate new data samples based on training data. They learn the underlying patterns and
Generative Models
structures of the data and can generate new content that resembles the training examples, enabling creative and generative applications.
Generative model types refer to specific architectures or algorithms used in generative AI. Examples include GPT (Generative Pre-trained Transformer), BERT
Generative Model
(Bidirectional Encoder Representations from Transformers), and GLaM (Generative Latent Algorithmic Models), each with unique characteristics and
Type
applications.
Deployment Deployment tools and technologies facilitate the deployment and hosting of generative AI models and applications. They provide functionalities for packaging
Tools/Technologies models, creating APIs, managing infrastructure, and deploying the solutions on servers or cloud platforms for accessible and scalable deployment.
Clouds refer to cloud computing platforms that provide on-demand access to computing resources and services over the internet. They offer scalability,
Clouds
flexibility, and cost-efficiency for hosting generative AI applications, enabling easy deployment and access to computational resources.
Machine Learning Machine learning frameworks are comprehensive software libraries that provide tools and APIs for developing and training machine learning models. They
Frameworks offer functionalities for tasks like data preprocessing, model building, and optimization.
Programming Generative AI programming languages are programming languages commonly used for implementing and executing generative AI algorithms. They provide
Languages libraries, frameworks, and syntax that enable developers to create and manipulate generative models.

Source: nasscom analysis

41 Generative AI Startup Landscape in India: A 2023 Snapshot

Authors

Namita Jain Dhiraj Sharma

AI Research Specialist, nasscom
Practice Lead, nasscom Insights Insights

Related nasscom Reports

Source: nasscom analysis

42 Generative AI Startup Landscape in India: A 2023 Snapshot

About nasscom
nasscom is the industry association for the technology sector in India. As a not-for-profit
organization funded by the industry, its objective is to build a growth-led and sustainable
technology and business services sector in the country with over 3100 members. nasscom
Insights is the in-house research and analytics arm of nasscom, generating insights and
driving thought leadership for today’s business leaders and entrepreneurs to strengthen
India’s position as a hub for digital technologies and innovation.

Plot 7 to 10, Sector 126, Noida – 201303, India

Phone: +91-120-4990111
Email: [email protected]

www.nasscom.in
Visit our community at
www.community.nasscom.in
43 Generative AI Startup Landscape in India: A 2023 Snapshot

Disclaimer
The information contained herein has been obtained from sources believed to be reliable. nasscom and its advisors & service providers disclaims all
warranties as to the accuracy, completeness or adequacy of such information. nasscom and its advisors & service providers shall have no liability
for errors, omissions or inadequacies in the information contained herein, or for interpretations thereof. The material or information is not intended
to be relied upon as the sole basis for any decision which may affect any business. Before making any decision or taking any action that might
affect anybody’s personal finances or business, they should consult a qualified professional adviser.

Use or reference of companies/third parties in the report is merely for the purpose of exemplifying the trends in the industry and that no bias is
intended towards any company. This report does not purport to represent the views of the companies mentioned in the report. Reference herein to
any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply
its endorsement, recommendation, or favouring by nasscom or any agency thereof or its contractors or subcontractors.

The material in this publication is copyrighted. No part of this report can be reproduced either on paper or electronic media without permission in
writing from nasscom. Request for permission to reproduce any part of the report may be sent to nasscom.

Usage of Information:
Forwarding/copy/using in publications without approval from nasscom will be considered as infringement of intellectual property rights.
Copyright © 2023

Plot 7 to 10, Sector 126, Noida – 201303, India

Phone: +91-120-4990111 | Email:[email protected]

www.nasscom.in
Visit our community at
www.community.nasscom.in /nasscomOfficial /nasscom /nasscomVideos