India Is Obsessed With Its Past. Who Is Building Its Future?

A decade ago, the Indian startup ecosystem was being celebrated as the next big thing.

Every other day, we heard about a new unicorn, a massive funding round, a founder disrupting an industry or a company preparing to change the world.

Entrepreneurship became fashionable. Startup founders became celebrities. Valuations became a measure of achievement.

But look at the ecosystem today.

Some founders who once spent their time talking about building revolutionary businesses are now spending a significant amount of time building their personal brands, making reels, selling courses and becoming influencers.

There is absolutely nothing wrong with content creation or personal branding. Founders are entitled to explore new opportunities.

But somewhere along the way, I feel we have started rewarding the ability to talk about building businesses more than the ability to build genuinely transformative businesses.

The incentives explain a lot. A podcast produces attention the same week; a hard technical problem may produce nothing visible for years. When applause arrives faster for storytelling than for engineering, people rationally drift towards storytelling. I have written before about how founders who perform success often confuse visibility with progress, and the pattern has only become more pronounced.

The money tells a similar story. WIPO's Global Innovation Tracker for India, published alongside the 2025 index, shows that the number of venture capital deals in India was flat between 2023 and 2024 and declined at an average of 1.5% a year between 2020 and 2024. Capital became scarcer and more cautious just as the conversation about founders became louder.

Where are the companies creating breakthrough technologies? Where are the startups developing products that the rest of the world cannot function without?

We have plenty of funding announcements, pitch decks, podcasts and success stories.

But where are the breakthroughs?

Walk into a supermarket and look at the products around you.

New flavours. New packaging. New variants. New combinations. New marketing campaigns.

The same categories, repackaged and relaunched.

Cold drinks, noodles, biscuits, ice creams, burgers and packaged food products dominate much of the consumer conversation.

There is nothing inherently wrong with the FMCG industry. It serves millions of consumers, creates employment and generates enormous economic value.

But is this what industrial innovation should look like?

When was the last time a major Indian consumer company introduced a genuinely disruptive product that created an entirely new category or changed consumer behaviour globally?

We seem comfortable competing over market share in existing categories rather than investing aggressively in creating new ones.

To be fair, India has produced organisational innovations in consumer markets. The sachet made premium products affordable one use at a time, and Amul built a cooperative model that made millions of small dairy farmers owners of a national brand. But these are decades old, and they were innovations in distribution and ownership rather than in the technology of what we make.

And this problem is not limited to FMCG.

Look at the larger industrial landscape. Where are our globally competitive breakthroughs in advanced materials, industrial robotics, precision manufacturing, semiconductor equipment, biotechnology and next-generation energy technologies?

DST's own data shows how narrow industrial research still is. In its Research and Development Statistics at a Glance 2025–26, transportation and pharmaceuticals accounted for 32.6% and 21.4% of industrial R&D. More than half of it is concentrated in vehicles and medicines.

Manufacturing is not the same as owning the technology

We talk about becoming a manufacturing powerhouse. But manufacturing something designed and invented elsewhere is not the same as owning the technology behind it.

Semiconductors are the clearest example. Under the India Semiconductor Mission, the government had approved 12 manufacturing projects with cumulative investment of about ₹1.64 lakh crore by July 2026, according to India Briefing's tally of official approvals. Of these, three are fabrication units and nine are assembly, testing and packaging units. Micron's Sanand plant, inaugurated in 2026, assembles and tests chips; it does not make wafers. A second phase, Semicon 2.0, approved in July 2026 with about ₹1.275 lakh crore, widens the focus to materials, equipment and design.

This is real, overdue progress. But packaging and testing are the least IP-intensive steps in the chip value chain, and the design tools, lithography machines and process know-how still belong to a few companies abroad. A plant on Indian soil running someone else's process is a factory. It is not yet a technology.

A country cannot become an innovation leader merely by becoming a bigger factory for the world.

This is perhaps the question that bothers me the most.

Companies like TCS, Infosys and Wipro have built enormous organisations. They employ hundreds of thousands of professionals, operate across continents and generate billions of dollars in revenue.

They are undoubtedly major achievements of Indian enterprise.

NASSCOM's Strategic Review 2026 estimated India's technology industry revenue at about $315 billion for FY26, with about 59.5 lakh people directly employed.

But look at how that revenue is made. By NASSCOM's own segment estimates, core IT services brought in about $149 billion in FY26, engineering and R&D services about $63 billion and business process management about $59 billion. Software products, the category in which a company owns what it sells, accounted for about $23 billion. Roughly one dollar in fourteen.

But let us ask a different question.

What globally dominant software product has emerged from these companies that has fundamentally changed the way the world operates?

Where is India's equivalent of Microsoft Windows, Google Search, AWS, Salesforce or NVIDIA's CUDA ecosystem?

Yes, these companies have developed platforms, enterprise products, intellectual property and substantial technology capabilities. Infosys Finacle, a core banking suite, and TCS BaNCS, a platform for banking, insurance and capital markets, are established enterprise software products used by financial institutions in many countries.

But the larger question remains: why haven't we produced more globally dominant technology platforms from organisations with such enormous talent pools, resources and customer access?

Part of the answer is the business model. A services company is paid for effort, people multiplied by hours, with predictable margins and growth that comes from hiring. A product company bets on outcomes years in advance, often while losing money. NVIDIA spent more than a decade building CUDA before the market understood what it was for. SAP built software that quietly runs the back offices of the world by investing in a product that its founders believed in long before it paid off. Those are not services decisions. They are research decisions.

Corporate India's thin R&D footprint

The European Commission's EU Industrial R&D Investment Scoreboard 2025, which ranks the world's 2,000 largest corporate R&D investors, is a useful mirror. It lists 674 companies from the United States, 525 from China, 318 from the EU and 192 from Japan. According to an analysis of the same data by Amit Kapoor and Sheen Zutshi of the Institute for Competitiveness, only 17 Indian companies make the list, together investing about €6.4 billion, roughly 0.44% of the total. Tata Motors alone, ranked 63rd globally, accounts for more than half of that.

Corporate research budgets decide which countries own the next generation of technology. On this list, India barely registers.

Source: European Commission JRC, EU Industrial R&D Investment Scoreboard 2025; India count via Institute for Competitiveness analysis of the Scoreboard (January 2026)

View the data

United States	674
China	525
European Union	318
Japan	192
India	17

Seventeen companies, in one of the five largest economies in the world. That number says more about our corporate priorities than any speech.

Where India is genuinely inventing

It would be unfair, and inaccurate, to say India invents nothing. My argument is that we do not invent enough, not that we invent nothing. The counterexamples matter, because they show exactly what is possible.

UPI and India Stack. The Unified Payments Interface is a genuine Indian innovation in financial infrastructure, and arguably the most important one of this century. In October 2026 it processed a record 25.47 billion transactions worth ₹31.53 lakh crore, according to NPCI data. ACI Worldwide's 2024 Prime Time for Real-Time report estimated that India accounted for 49% of all real-time payment transactions worldwide in 2023. It is now live in 11 foreign countries, including the UAE, France and Singapore. This is precisely the kind of thing I am asking for: a technology designed in India that other countries now adopt.

ISRO. On 23 August 2023, Chandrayaan-3's Vikram lander touched down in the Moon's south polar region, making India the fourth country to soft-land on the Moon and the first to do so near its south pole. ISRO shows what Indian engineers achieve when an institution is given a long-term mission and the patience to pursue it.

Global software products. Zoho, built by Sridhar Vembu without venture capital, sells business software to customers worldwide from offices that include rural Tamil Nadu. FusionCharts was bootstrapped into a global charting product from Kolkata by a teenager. Freshworks listed on Nasdaq in 2021. These companies prove that the "India cannot build products" story is false.

Pharmaceuticals and vaccines. India's generic drug industry supplies affordable medicines to much of the developing world, and Covaxin, developed by Bharat Biotech with the Indian Council of Medical Research, showed that indigenous vaccine development is possible at speed.

But look at the list again: a not-for-profit payments network, a government space agency and a few founder-led companies that ignored the dominant model around them. Very few came from our largest private corporations, and that is the point.

And now comes AI

We have millions of engineers, enormous IT services organisations and decades of experience delivering technology solutions to global businesses.

Yet, when the world is competing to build foundational AI models, advanced computing infrastructure, semiconductor ecosystems and next-generation platforms, where is India's equivalent contribution at comparable global scale?

The government has started to act. The IndiaAI Mission was approved in March 2024 with an outlay of ₹10,371.92 crore over five years. By February 2026, according to a Press Information Bureau backgrounder, more than 38,000 high-end GPUs had been onboarded and made available at about ₹65 per hour, and 12 startups had been selected from more than 500 proposals to build foundation models, including Sarvam AI, Soket AI and Gnani AI, alongside the IIT Bombay-led BharatGen consortium. Several of these teams unveiled models at the India AI Impact Summit in New Delhi in February 2026.

I welcome every one of these efforts. But the compute pooled for the entire national mission is modest next to the private clusters of the largest American and Chinese technology companies, the same companies that dominate the R&D scoreboard above. India's largest technology firms have so far focused mainly on implementing AI for clients rather than building foundational models.

Indian companies are certainly participating in AI through research, platforms, services and enterprise implementations. But participating in a technological revolution and leading one are two very different things.

We cannot keep measuring our technological strength through the number of engineers we employ or the number of projects we deliver for foreign companies.

At some point, we must start measuring it through the technology we own.

India has an extraordinary talent pool.

Our engineers work at the world's largest technology companies. Indian-origin professionals lead some of the biggest global organisations. Our scientists, researchers and entrepreneurs have demonstrated what they are capable of when given the right environment.

So the problem cannot simply be a shortage of intelligence or talent.

Perhaps we have built an ecosystem that rewards execution more consistently than experimentation.

Our businesses often prefer predictable returns over uncertain research. Investors want quicker exits. Founders are encouraged to demonstrate traction before they have developed something genuinely differentiated.

Companies want experienced employees who can deliver projects immediately, but how many are willing to invest patiently in research that might take five or ten years to produce a breakthrough?

Who actually does India's research?

The data answers this question more clearly than any opinion. DST's latest survey shows that in 2023–24, business enterprises (public and private industry together) performed 49.3% of India's R&D, government institutions 34.9%, higher education institutions 12.6% and scientific and industrial research organisations 3.2%.

Businesses now carry about half of India's research effort. In the United States the business share is 78% (NSF, 2023).

Source: Department of Science and Technology, Research and Development Statistics at a Glance 2025–26 (NSTMIS, 2026)

View the data

Business firms	49.3%
Government bodies	34.9%
Higher education	12.6%
Research orgs	3.2%

That business share is a genuine improvement. DST reports that private industry's share rose from 32.2% in 2020–21 to 45.2% in 2023–24. But compare it with the economies we want to catch. In the United States, the business sector performed 78% of all R&D in 2023, according to the National Science Foundation. In the European Union the figure was 66%, according to Eurostat. In innovation leaders, companies, not governments, carry most of the research burden, because they are the ones who turn research into products.

A second detail deserves attention. Multinational enterprises accounted for 71.3% of business R&D spending in India in 2023–24, up from 49.9% in 2020–21. The category includes Indian multinationals as well as foreign ones, so it is not a measure of foreign dominance. But it suggests that much of the growth in industrial research comes from large global groups, some of whose intellectual property is owned and commercialised outside India. India is increasingly where research is done. It is not yet, to the same degree, where research is owned.

Compare this with countries that made research a national habit. Israel built a culture in which engineers leave military technology units to found companies like Check Point, and the World Bank's data shows it spending more than 6% of GDP on R&D in 2023. Britain produced James Dyson, who built thousands of prototypes before a single product worked. These are not stories about talent. India has talent in abundance. They are stories about patience and about who is prepared to pay for failure.

Innovation requires the freedom to fail, capital that can tolerate uncertainty, institutions that encourage experimentation and businesses willing to think beyond the next quarterly result.

We cannot build the next generation of global technology companies with the same mindset we use to run an outsourcing business.

This is not just a feeling. Some of the numbers deserve our attention.

of GDP spent on research and development, 2023–24

0.84%

DST, R&D Statistics at a Glance 2025–26

of 139 economies in the Global Innovation Index 2025, unchanged in 2026

38th

WIPO

Indian companies among the world's top 2,000 corporate R&D investors

17

EU Industrial R&D Investment Scoreboard 2025

The innovation ranking

According to the World Intellectual Property Organization's Global Innovation Index 2025, India ranked 38th among 139 economies. In the 2026 edition, released in Geneva on 29 September 2026, India held the same position.

That is progress. India has improved significantly from its 81st position in 2015.

The climb from 81st to 38th is real, but the line has been almost flat since 2022.

Source: WIPO, Global Innovation Index India country profiles (2018, 2025, 2026 editions)

View the data

2015	81
2016	66
2017	60
2018	57
2019	52
2020	48
2021	46
2022	40
2023	40
2024	39
2025	38
2026	38

The climb is real and deserves credit. India has been named an innovation "overperformer" relative to its income level for 16 consecutive years, and it ranks first among lower-middle-income economies. But the curve has flattened since 2022, and WIPO's 2026 profile for India is telling: India ranks 25th for innovation outputs but only 49th for innovation inputs, which include research spending, human capital and business sophistication. We are producing more than our investment would predict. That is impressive, but it cannot last without more investment.

But compare that with China at 10th place. China entered the top 10 for the first time in 2025 and held 10th in 2026, the only middle-income economy among the innovation leaders.

How much we spend on research

More importantly, India's gross expenditure on research and development was approximately 0.84% of GDP in 2023–24, according to the Department of Science and Technology.

That figure comes from DST's Research and Development Statistics at a Glance 2025–26, based on the National S&T Survey 2024–25 and released in mid-2026. In rupee terms, GERD reached about ₹2.45 lakh crore in 2023–24, up from about ₹79,356 crore in 2013–14. The ratio is up from 0.64% in 2020–21, which was the figure in the previous edition of the same publication.

Two cautions. Part of the jump reflects better measurement, because the new survey collected spending directly from institutions, including universities, and analysts note the data cannot fully separate real growth from improved coverage. And even at 0.84%, India remains far behind its peers. DST itself recommends raising GERD above 1% of GDP in the near term and towards 2% over the long term.

For a fair comparison, the chart below uses one source and one year for everyone: 2020, the latest year with World Bank (UNESCO) figures for all these economies, including India.

In 2020 Israel spent about nine times India's share of GDP on research, South Korea about seven times and China more than three times.

Source: World Bank, World Development Indicators, GB.XPD.RSDV.GD.ZS (UNESCO Institute for Statistics data), accessed November 2026

View the data

Israel	5.83%
South Korea	4.52%
United States	3.42%
Japan	3.27%
Germany	3.09%
China	2.36%
Brazil	1.22%
India	0.65%

The gap has not closed since. For 2023, the same World Bank series shows Israel at 6.35%, South Korea at 4.94%, the United States at 3.45%, Japan at 3.44%, Germany at 3.15% and China at 2.58%. Set DST's 0.84% for 2023–24 against those numbers and the conclusion is unavoidable: China spends roughly three times as much of its GDP on research as we do, on an economy several times larger. In absolute terms the difference is not a gap. It is a chasm.

Patents: more filings, not yet more inventions

There is a more encouraging number, and it needs careful reading. Patent filings at the Indian Patent Office have nearly doubled in four years.

Filings almost doubled in four years. More filings are not yet the same as more inventions that reach the market.

Source: Office of the Controller General of Patents, Designs and Trade Marks, Annual Report 2024–25

View the data

2020–21	58,503
2021–22	66,440
2022–23	82,811
2023–24	92,168
2024–25	1,10,375

In 2024–25, 110,375 applications were filed, and Indian residents accounted for 68,201 of them, or 61.79%, up from 55.96% the year before. For a country that long left most of its patenting to foreign applicants, that shift is significant. WIPO's World Intellectual Property Indicators 2025 placed India sixth in the world for patent filings.

But a filing is not an invention, and an invention is not a product. The economist Sunil Mani, writing in The India Forum in March 2026, points out that applications from educational institutions rose from 11,150 in 2020–21 to 45,776 in 2024–25, driven partly by university ranking metrics (NIRF) that award points for filing whether or not a patent is ever granted or used. Only 33,504 patents were granted in 2024–25. The share of granted patents that are reported as commercially worked, he notes, remains in single digits. Patent counts can be gamed. Products in the hands of customers cannot.

We are also publishing more science: DST counts 2,07,390 Indian publications in 2022, third in the world, and about 354 full-time-equivalent researchers per million people. But publications, like patents, matter only when someone turns them into something useful.

For a country aspiring to become a developed economy, this raises a serious question about the scale of our commitment to research.

We cannot expect world-changing inventions while treating research as an expense that can always be postponed.

Our ambition cannot be several times larger than our investment in discovering what comes next.

India has a tremendous opportunity.

We have a huge domestic market, a young workforce, a growing digital economy, expanding infrastructure and a rapidly developing startup ecosystem.

But economic growth and technological leadership are not interchangeable.

A country can grow its GDP through consumption, services, construction, trade and manufacturing.

But becoming a technological leader requires something more.

It requires owning intellectual property, creating original products, developing advanced manufacturing capabilities, commercialising scientific research and building companies that establish entirely new global markets.

Consider how value is distributed in a modern technology product. The owner of the design, the software and the patents captures most of the margin. The assembler captures a sliver. The country that supplies the engineers captures their salaries, but not the platform or the royalties that compound for decades. India is well positioned in the middle of that chain and weakly positioned at the top.

We should not be satisfied with being the country that supplies the world's engineers while someone else owns the platforms they build.

We should not be satisfied with assembling products when the most valuable technologies, designs and patents belong elsewhere.

And we should not confuse a billion-dollar valuation with a billion-dollar contribution to technological progress.

Imagine if even a fraction of the energy we spend celebrating startup valuations and corporate achievements went into building the following:

• Artificial intelligence: Foundational models, computing infrastructure, specialised AI chips and original AI research.
• Advanced manufacturing: Industrial robotics, precision machinery, semiconductor equipment and automation systems.
• Healthcare: Original medical devices, biotechnology, affordable diagnostics and breakthrough pharmaceutical research.
• Energy: Battery technology, energy storage, advanced solar technology and next-generation nuclear systems.
• Agriculture: Precision farming, agricultural robotics, food processing technology and scientific solutions that improve productivity.
• Software: Globally competitive SaaS platforms, developer tools, cybersecurity products and enterprise technology that we own and export.

We don't need to invent everything. No country does.

But we need to move from being predominantly consumers, implementers and service providers of other people's innovations to becoming creators and owners of important technologies ourselves.

What it would take

India has finally built some of the right policy instruments. The question is whether they are big enough to change behaviour.

Patient public capital for private research. The Union Cabinet approved the Research, Development and Innovation (RDI) scheme in July 2025 with a corpus of ₹1 lakh crore, designed to provide long-term, low-cost or zero-interest financing for private R&D in strategic and sunrise sectors. Launched in November 2025, its fund sits within the ANRF and invests through second-level fund managers. On 8 October 2026 the first cohort of 19 managers (14 alternative investment funds and 5 focused research organisations) was approved, with an RDIF commitment of ₹7,131 crore, which they are expected to match with private capital. This is exactly the patient, risk-tolerant money our ecosystem has lacked. The measure of success should be the number of products it produces, not the volume of funds it deploys.

A real national research foundation. The ANRF, created by the Anusandhan National Research Foundation Act, 2023, was approved with an estimated budget of ₹50,000 crore over 2023–28, of which about ₹36,000 crore was expected to come from non-government sources such as industry and foundations. That design is ambitious, and the private contribution is where it will succeed or fail. If Indian companies treat the ANRF as a government programme rather than a partnership, the money will not arrive.

Corporate R&D that is measured and rewarded. The most powerful lever is not in Delhi but in boardrooms. Large listed companies could report R&D intensity alongside revenue growth, protect multi-year research budgets from bad quarters and tie leadership incentives to products launched and IP commercialised. Analysts could ask about R&D on earnings calls as routinely as they ask about margins.

Fixing the incentives in universities. If ranking frameworks reward patent filings regardless of grants or commercial use, we will get filings. As Mani argues, metrics that weight granted, licensed and commercialised patents would reward real research rather than paperwork.

Government as first customer, and investors who can wait. Many great technologies were bought by governments before anyone else would buy them; public institutions can be early adopters of indigenous designs, as the state was when it backed UPI. And capital built for consumer apps expects exits in five to seven years, while semiconductor equipment, biotechnology and energy storage need a decade or more.

None of these ideas is radical. Every country that leads in technology has done some version of all of them.

We often talk about Viksit Bharat 2047.

It is an ambitious national aspiration, and there is nothing wrong with having a big national goal. The government's stated vision is for India to become a developed nation by 2047, the hundredth year of independence, and even the RDI fund's own announcement frames its purpose as transforming India "from a technology consumer" into a nation that develops globally competitive technologies.

But a date on a calendar does not create a developed country.

The question is what we do between now and then.

Are our universities producing research that industries are willing to commercialise?

Are our large corporations investing in technologies that might not generate immediate returns?

Are our entrepreneurs solving difficult problems or simply identifying the next business model that can attract funding?

Are our investors willing to back deep technology companies for a decade?

Are we building products that the rest of the world wants to buy because nobody else can build them as well?

And most importantly, are we creating an environment where the next generation wants to become scientists, inventors, engineers and builders rather than simply aspiring to become the next unicorn founder or social media celebrity?

These are the questions that matter.

There are twenty-one years left. That sounds like a long time. In research, it is not. A breakthrough battery chemistry, a new class of drug or a domestic chip-making tool can take a decade or more to move from laboratory to market. The research that will define India in 2047 has to be funded now, by people who will not be in charge when it pays off.