In today's deep dive, we will examine how India's growing GCC ecosystem is changing the startup landscape. As multinationals expand into AI, product development and R&D, they are competing with startups for talent, customers and expertise. Yet these centres could also become customers, partners, acquirers and sources of future founders. We will explore both sides, the impact on startup economics and innovation, and who ultimately benefits from the technology being built in India.

In March 2025, JPMorganChase brought its global software engineering conference to Hyderabad for the first time. Around 350 developers, data scientists and technology specialists gathered to discuss the bank's engineering work. What made the event interesting was where it was being held. About one-third of JPMorganChase's 63,000 technology professionals were already based in India, working on systems used by one of the world's largest financial institutions.

Now consider a Bengaluru startup trying to recruit an experienced AI engineer or cybersecurity specialist. It might have an interesting product, a decent funding round and stock options that could become valuable someday. But across the city, JPMorganChase can offer the same engineer an established global employer, competitive compensation and the opportunity to work on sophisticated financial technology. Goldman Sachs, Target, Walmart and dozens of other multinationals are competing for similar people.

For a long time, Indian startups mainly worried about competing with one another for talent. They also lost engineers to Google, Amazon and Microsoft, but those companies were relatively concentrated employers of premium technology talent. The expansion of Global Capability Centres, or GCCs, is bringing many more multinational employers into that competition.

And hiring is only part of the story. These centres increasingly build software products, develop AI systems and take responsibility for research that was previously handled at overseas headquarters. Some are doing exactly the kind of work Indian startups hope to sell to large enterprises.

India has spent years celebrating the growth of both its startup ecosystem and its technology services industry. GCCs are now bringing these two worlds much closer together, creating opportunities for some startups and serious problems for others. The interesting question is no longer how many GCCs India can attract, but what their expansion will mean for the Indian technology companies trying to grow alongside them.

The GCC story started with a relatively simple economic argument. A multinational could move finance, customer support, technology maintenance and other business operations to India at a lower cost. The work remained within the company's own organisation, unlike traditional outsourcing, but India offered substantial savings.

Over time, these centres became more sophisticated. Companies discovered that their Indian teams could do much more than maintain existing systems. They could develop products, manage global technology platforms and undertake specialised research. Senior engineering roles and decision-making responsibilities gradually moved to India as well.

The numbers show how far this has progressed. According to the Nasscom-Zinnov GCC Landscape Report 2026, India now hosts 2,117 GCCs operating across 3,728 units. They employ 2.36 million people and generate approximately $98.4 billion in annual revenue. The number of centres has grown 32% since FY21, and 506 companies from the Forbes Global 2000 have a GCC presence in India.

The nature of their work is changing just as rapidly. Zinnov classifies 39% of Indian GCCs as portfolio hubs, which take responsibility for entire products, platforms or intellectual property. Another 5% operate as transformation hubs with broader responsibility for technology and business change. The remaining 56% largely focus on delivery and functional operations, so it would be wrong to describe every GCC as an advanced research centre.

Nevertheless, a substantial part of the industry has moved beyond the model that originally brought it to India. Zinnov's research also indicates that around 45% of the work performed across Indian GCCs now involves specialised expertise or advanced technical capabilities.

Consider what companies are building here. Boeing invested ₹1,600 crore in a 43-acre engineering and technology campus in Bengaluru, its largest such investment outside the United States. The facility includes advanced research and testing capabilities. In March 2026, Airbus inaugurated an 880,000-square-foot technology centre in Bengaluru designed to accommodate approximately 5,000 employees. Its work extends across aircraft engineering, digital systems, cybersecurity, robotics and research.

Target offers another example. Around 5,600 people worked at its India centre in May 2026, including roughly 40% of the retailer's global technology workforce. Indian teams support digital commerce, merchandising, supply chains and analytics.

These are not small support offices helping overseas headquarters with routine work. They are becoming part of how major global businesses develop and operate their technology.

That changes the competition facing Indian startups.

The first problem is the growing cost of hiring experienced technical people.

Startups have traditionally used a combination of salaries, interesting work, faster promotions and employee stock options to attract engineers. For a young developer, joining a promising startup could mean learning quickly and working directly with its founders. If the company succeeded, ESOPs offered the possibility of a significant financial return.

GCCs increasingly offer much of the professional appeal without requiring employees to take the same financial risk. An engineer can work on advanced AI, financial trading systems, aerospace software or cybersecurity while receiving compensation supported by a multinational's established business. Some GCCs also offer equity-linked incentives and international career opportunities.

Zinnov's 2026 salary survey, covering more than 95 GCCs and 150 centres, reported an average salary increase of 9.8%. But the average hides substantial differences. AI and machine learning roles recorded increases of 21.1%, while cybersecurity roles saw 20% and cloud roles 16.1%. Attrition among high-performing employees was 16.5%.

These figures measure salary increases within GCCs, not the exact premium they pay over startups. Nevertheless, they illustrate where competition is most intense. Employers are becoming more selective about the skills they pay heavily for, and experienced AI engineers, security specialists and engineering architects are among the beneficiaries.

For a startup, even a modest change in salary expectations can become expensive. Consider a Series A software company employing 20 engineers at an average annual employment cost of ₹30 lakh. Its engineering payroll is already ₹6 crore. If competition pushes that cost up by 20%, the annual bill increases to ₹7.2 crore.

The additional ₹1.2 crore does not necessarily produce more customers or revenue. It may simply be the cost of retaining the existing team.

The damage can extend further when a critical employee leaves. Replacing a senior engineer takes time, and the new employee needs to understand the product before becoming fully productive. Losing the person responsible for a major customer's implementation could delay the project by months.

A multinational with thousands of engineers can often move people between teams. A startup with 25 employees may have only one person who fully understands a particular part of its technology.

There is an important distinction here, however. GCCs are not taking engineers exclusively from startups. Recruitment firm Xpheno estimated in 2025 that 45% to 48% of GCC hiring came from traditional IT services companies. Large firms such as TCS, Infosys, Wipro and other technology employers are an important part of this talent market.

Well-funded startups can also compete effectively, and some pay considerably more than GCCs. The problem is more acute for companies that need experienced engineers but have limited financial resources.

That places many Series A and Series B startups in a difficult position. Seed-stage companies can sometimes persuade engineers to join for the experience and potential upside. Large startups may have the money to match competing offers. Companies in between need strong teams to build reliable products, but often cannot afford to keep increasing compensation.

The competition may be particularly damaging in deep tech, where replacing a specialised researcher or engineer is much harder than replacing someone working on a relatively common software application.

This competition is arriving at a time when venture capital has become much more selective.

According to Tracxn, Indian technology startups raised $10.3 billion during the first nine months of 2026, an increase of 7% over the corresponding period last year. On the surface, funding appears to be recovering.

The number of deals tells a different story. Funding rounds declined 38% to 1,134. Seed funding fell 37% to $698 million, while the number of startups receiving their first funding round dropped 30% to 338.

Capital is still available, but investors are concentrating more of it in fewer companies. This makes the market especially difficult for founders who have not yet demonstrated strong customer demand.

GCCs operate under a different financial model. An established multinational can fund a technology project using its operating budget, provided the project makes economic sense for its existing business. The team does not need to raise another funding round every 18 months or convince a venture capitalist that its product could eventually become a large independent company.

Of course, multinational budgets are not unlimited. Projects get cancelled, hiring freezes happen and corporate restructuring can eliminate entire departments. GCCs face their own pressure to justify costs.

But the financial starting point is still different.

Imagine two teams building similar fraud-detection technology. One is an independent startup developing a product for banks. The other works inside the Indian technology organisation of an international bank.

The startup needs to finance product development, obtain regulatory and security approvals, acquire customers and demonstrate that other banks will pay for its software.

The internal team may only need to show that its product reduces fraud losses or improves efficiency within the parent bank. It already has access to relevant business teams, operational information and a known customer.

The startup might ultimately develop a better product because it learns from multiple banks and has the freedom to sell widely. But reaching that point requires substantially more commercial effort.

It would be misleading to compare the GCC industry's $98.4 billion annual revenue with Indian startup venture funding and conclude that GCCs have more investment capital. The two figures measure different things. GCC revenue largely reflects services and operations performed within multinational groups, whereas venture funding is capital invested in businesses.

What matters is that a GCC-backed project can be supported by an existing enterprise's financial resources, while an independent startup must establish a business capable of supporting itself.

As venture investors demand faster progress towards profitability, this difference becomes harder to ignore.

The bigger threat to startups may not be the engineers GCCs employ, but what those engineers are capable of building.

Enterprise software has historically been an attractive startup category. Large organisations have complex operations, and independent software companies can develop products that solve specific problems more efficiently than traditional manual processes.

But many multinationals now have extensive engineering teams in India working on those same problems.

Target's Indian employees work on retail analytics, digital commerce and supply-chain technology. Goldman Sachs develops sophisticated financial and risk-management systems through its Indian technology operations. Airbus engineers work on aircraft systems and digital capabilities.

These companies can still purchase external software, but their ability to develop alternatives internally has improved significantly.

Consider a startup selling inventory optimisation software to large retailers. Its product analyses demand, recommends stock levels and helps minimise excess inventory. For a retailer without a sizeable technology team, buying the software may be the obvious choice.

Now consider a retailer with hundreds of engineers, years of sales information and an established data infrastructure. Building an internal system becomes a more realistic option, particularly if the retailer believes the technology provides a competitive advantage.

The startup is no longer competing only with other software vendors. It is competing with the customer's own engineering department.

The arrival of generative AI makes this calculation more interesting.

Building useful software still requires experienced engineers, reliable data, security and ongoing maintenance. AI does not eliminate those requirements. But it can reduce the time needed to develop certain applications, especially relatively simple internal tools and repetitive workflows.

A multinational that previously needed an external vendor for process automation may now be able to build a satisfactory version internally with a smaller engineering team.

Zinnov's March 2026 study examined 1.7 million job descriptions across more than 200 enterprises and estimated that around 55% of India's GCC work portfolio was exposed to potential displacement by AI. That figure should not be interpreted as 55% of employees losing their jobs. It measures the kinds of work that automation could affect.

There is a second consequence that receives less attention. If AI enables GCCs to employ fewer people for routine tasks, they may redirect spending towards smaller teams of expensive specialists. Such organisations could become more capable of building internal products even without substantial increases in headcount.

For startups selling basic workflow automation or generic analytics, this is a real concern. They will need to demonstrate an advantage beyond assembling existing AI models and connecting them to an enterprise's internal systems.

However, developing an internal tool is not the same as maintaining a commercial software product. The retailer may be capable of building inventory software but decide that its engineers are better employed improving customer experience or fulfilment systems. External vendors also spread development costs across many customers.

This is where the distinction between ordinary software and a durable software business becomes important. A startup with proprietary information, difficult integrations, specialised expertise or a product used successfully by many enterprises can remain attractive even to sophisticated technology buyers.

The danger is greatest when the customer's engineers can reproduce most of the startup's value within a few months.

There is another side to this relationship, and it could be particularly valuable for Indian enterprise startups.

For years, founders building products for American banks, retailers and industrial companies often needed to sell directly to decision-makers overseas. That meant expensive international travel, local sales teams, long procurement processes and substantial effort convincing unfamiliar organisations to trust a small Indian company.

Today, some of the people who evaluate these products are working in Bengaluru, Hyderabad and Pune.

JPMorganChase is a useful example. Its third Technology Innovation Forum in Bengaluru in 2025 invited more than 40 startups to present technology involving payments, cybersecurity, cloud engineering, data and developer tools. In May 2026, the bank held its fourth forum, continuing its engagement with startups, venture capital firms and research institutions.

A startup can now potentially build relationships with engineers and technology leaders employed by one of the world's largest banks without first establishing a major international sales presence.

That does not mean procurement decisions have become easy. A technology demonstration in Bengaluru still needs to pass security reviews, commercial evaluations and internal approvals. The authority to sign a global contract may remain at headquarters.

Nevertheless, easier access to knowledgeable enterprise buyers can reduce one of the largest difficulties faced by Indian business-to-business software companies.

Shell offers a more concrete example of how these relationships can develop.

Through its E4 programme, established in 2017, Shell has worked with more than 40 startups focused on energy, industrial technology and related areas. Participating companies can access technical expertise, research facilities, commercial pilots and potential financial support.

Shell's published examples include Detect Technologies, which developed technology for monitoring industrial assets and improving safety. Its technology was brought into Shell's industrial operations. Ossus Bio used Shell's laboratory infrastructure to pilot technology related to biological hydrogen production, while Energos applied AI tools to energy-efficiency problems at Shell retail sites.

These examples are particularly significant for deep-tech companies. A startup developing industrial sensors, advanced materials or energy systems cannot always validate its product inside an ordinary office. It needs specialised equipment, real operating environments and organisations willing to test technology that has not yet been widely commercialised.

Corporate research facilities can help provide those conditions.

Shell also expanded potential financial support for selected startups in its 2025 programme to between $100,000 and $500,000 through convertible loans, subject to evaluation and due diligence.

But founders should be careful not to confuse corporate interest with commercial success. A pilot can establish technical credibility without producing substantial revenue. Some corporations run startup programmes to understand emerging technologies, and the companies participating do not necessarily become long-term suppliers.

The stronger measure is whether a startup wins a paid deployment, whether the customer expands that deployment and whether the technology can subsequently be sold to other enterprises.

A startup with one successful multinational customer and the freedom to sell to its competitors could build a valuable business. A startup dependent on a single corporation for funding, technical access and most of its revenue could end up resembling a captive contractor.

The distinction may determine whether the partnership creates an independent company or simply provides another way for a multinational to obtain specialised development work.

The relationship could eventually extend from customers and partnerships to acquisitions.

Multinational companies regularly face decisions about whether to develop technology internally or purchase it from an outside business. When a startup has already solved a difficult problem, buying its technology and engineering team can be faster than building the same capability from scratch.

Lowe's acquisition of the retail analytics platform of Boomerang Commerce in 2019 provides an interesting precedent.

The US home-improvement retailer wanted to improve its pricing and merchandise decisions. Instead of relying entirely on internal development, it acquired Boomerang's analytics platform. The transaction included proprietary technology and employees in Bengaluru and the United States.

Importantly, Lowe's did not acquire the entire company. Boomerang's separate CommerceIQ business continued independently.

This is a useful example of how a large enterprise can acquire a particular technology asset and some of the people behind it without purchasing every part of the original startup.

For Indian startups, such transactions could become an alternative exit route. A company may struggle to build a large independent business but develop valuable technology that a multinational wants to integrate into its existing operations.

Indian GCC executives could play an important role in identifying such companies, testing their products and recommending acquisitions to overseas headquarters.

However, there is not yet enough public deal-level evidence to claim that GCC-driven acquisitions have become a major exit category for Indian startups. The possibility is worth investigating, but it should not be treated as an established trend.

There is also a risk that the growing ability of GCCs to recruit experienced engineers reduces their willingness to acquire businesses simply for their teams. A multinational may prefer hiring five specialists rather than purchasing a startup with 30 employees, existing liabilities and uncertain commercial prospects.

That means acquisition prospects will depend on what the startup actually owns. Proprietary technology, hard-to-obtain customer relationships and products that would take years to reproduce are more compelling reasons to acquire a business than the presence of a talented engineering team alone.

There is a larger economic question behind all this growth.

Suppose 1,000 engineers in Bengaluru develop a sophisticated financial software platform for a global bank. They may be responsible for the architecture, development, testing and continuing improvement of that product.

The work happens in India, and Indian professionals receive the salaries and experience associated with it. But ownership of the underlying technology may belong to the multinational parent company or another entity within its corporate group.

If that technology eventually helps the bank earn billions of dollars, the commercial benefits primarily accrue to the business that owns and deploys it.

Now consider an independent Indian startup developing a successful financial software platform. The company owns or controls the technology it creates, subject to its contracts, and can potentially sell it to customers around the world. If it becomes a valuable business, that value accrues to its shareholders, including founders, investors and employees holding equity.

These are different ways of participating in the global technology economy.

GCCs bring substantial benefits through jobs, professional training, local expenditure, taxes and participation in global research. There is no reason to dismiss those benefits simply because their parent companies are based overseas.

But creating technology and owning the business that commercialises it are not the same thing.

India's IT services industry demonstrated that the country could supply technology expertise at enormous scale. The startup ecosystem developed around a different ambition: creating companies capable of owning products, controlling distribution and capturing more of the profits generated by technology.

GCCs sit somewhere between these models. Their employees may have greater responsibility for important products than traditional outsourced teams, while commercial ownership generally remains with the parent group.

The long-term economic outcome depends partly on whether the expertise developed inside these centres eventually spreads into independent Indian businesses.

That brings us to another possibility. GCCs could become one of India's most important sources of future technology founders.

There is already a history of entrepreneurs emerging from multinational technology organisations in India.

Sachin Bansal and Binny Bansal worked at Amazon's India Development Centre before establishing Flipkart in 2007. Their time inside a global technology company gave them experience of building and operating internet products, although Flipkart's success required developing a business suited to India's very different retail environment.

Another relevant example is Geetha Manjunath, the founder of Niramai, which developed technology for non-invasive breast cancer screening. Before starting the company, she headed research at Xerox in India and worked as a principal research scientist at HP Labs.

Her career illustrates how experience in advanced corporate research can become the foundation for an independent technology business.

As GCCs take responsibility for more complex global products, they could help develop a larger pool of engineers who understand sophisticated technology, large enterprise problems and international customers.

That experience is valuable for entrepreneurship. A founder who has spent years developing fraud-detection systems inside a major bank might understand the limitations of existing products better than someone approaching the sector without industry experience.

But there is a competing force.

An engineer earning a substantial salary at a multinational, receiving equity incentives and enjoying opportunities to work abroad may have less financial motivation to leave and start a company. Entrepreneurship becomes particularly difficult when the person has a family, financial commitments and limited savings.

The GCC could therefore produce more people capable of becoming founders while making the decision to become one less attractive.

There are also practical restrictions. Employees cannot simply take proprietary data, confidential technology or intellectual property developed for their employers into new businesses. Founders emerging from GCCs must build their companies around knowledge and experience they are legally free to use.

The most interesting outcome may be the emergence of specialised startups founded by people who have spent years understanding difficult industry problems. They may be fewer in number than generic software ventures, but potentially much stronger businesses.

State governments have become enthusiastic supporters of the GCC industry because of its ability to bring high-paying jobs, global investment and demand for local infrastructure.

Maharashtra's GCC Policy 2025 aims to attract around 400 new centres and create four lakh high-skilled jobs. It offers eligible companies incentives including capital subsidies, rental assistance, payroll support and research grants.

For certain large GCC projects, capital subsidies can reach ₹100 crore, subject to eligibility and government approval. The policy also provides payroll support calculated at 40% or 50% of the eligible salary component above ₹1 lakh per month for specified employees, subject to limits and conditions.

These are policy provisions, not evidence that every eligible company has received the full benefit.

Karnataka has also developed a GCC policy covering 2024 to 2029, with ambitions to attract 500 additional centres and create 3.5 lakh jobs.

The logic behind these policies is understandable. Multinationals employ large numbers of people, occupy office space and create substantial local economic activity. A successful GCC can provide relatively visible employment and investment outcomes.

But governments are simultaneously using other programmes to encourage entrepreneurs to establish Indian-owned technology companies.

That creates an interesting policy question. If states are willing to support multinational technology employers financially, should they also expect measurable contributions to local startup ecosystems?

Such contributions could include purchasing technology from Indian companies, supporting university research that can be commercialised independently, opening specialised testing facilities to startups and helping local suppliers reach international customers.

Simply requiring corporations to conduct occasional startup events would accomplish little. The more meaningful measure would be whether those partnerships lead to commercial contracts, useful technology development and businesses that can survive independently.

The geographical impact also deserves attention. Bengaluru and Hyderabad are already expensive technology markets, and the concentration of GCCs adds to demand for experienced professionals and commercial infrastructure. Their expansion into Pune, Coimbatore, Jaipur, Kochi and other cities could create similar pressures, although these locations may benefit from lower costs and improved employment opportunities.

A smaller city could gain a much stronger engineering workforce because of GCC investment. Over time, that workforce might also support local startups. Whether the benefits outweigh the rising competition for talent will depend on how the local ecosystem develops.

The consequences will not be equal across the ecosystem.

Generic enterprise automation companies appear particularly exposed. If a startup's product mainly connects existing software systems and automates a relatively straightforward process, a capable GCC team may be able to build an adequate alternative.

The same applies to some analytics businesses and companies selling basic AI applications without much proprietary technology. Their customers may prefer internal development, especially when sensitive enterprise information is involved.

Custom software development firms face a related problem because GCCs can replace some work that would otherwise be outsourced to independent suppliers. The economics are different for technology requiring exceptional expertise, but routine development services are increasingly vulnerable to captive engineering teams.

At the other end, specialised cybersecurity, industrial software, semiconductor, medical technology and other deep-tech companies may find new customers and research partners among GCCs.

Building an advanced industrial sensor or a validated medical technology product is very different from assembling a simple software workflow. A multinational might possess the money and engineers to develop such technology, but purchasing a proven product can still save years of research and testing.

Enterprise software companies with substantial customer networks, difficult integrations and established products could also benefit. The more sophisticated GCCs become, the better equipped they may be to understand and purchase technically demanding products.

Consumer startups face a less direct effect. A quick-commerce company or consumer brand may compete with GCCs for certain engineers, but its core commercial challenges involve customers, distribution, operations and margins. These are not problems a multinational engineering centre automatically solves.

From an investment perspective, the most useful question may now be whether a startup's potential customers could build a satisfactory version of its product using their own Indian engineering teams.

If they can, the founder needs a convincing explanation for why buying the product is still the better decision.

The answer could involve cost, speed, specialist knowledge, accumulated data or the difficulty of maintaining software over many years. But simply having developed a working product may no longer be enough.

For founders, this also changes the importance of distribution. A technically superior product does not automatically win an enterprise customer. The company must reach decision-makers, negotiate contracts, support deployments and continue improving its offering. GCCs can become valuable routes to those customers, but only when they have meaningful purchasing authority or influence over global procurement.

The GCC boom is a substantial achievement for India. A country that once attracted multinational companies primarily because engineering labour was cheaper is now entrusted with parts of their most important products and research programmes.

Boeing and Airbus developing advanced technology in Bengaluru, JPMorganChase running global engineering activities from India and international retailers relying on Indian teams for their digital operations all demonstrate how far the industry has evolved.

The concerns for startups are equally real. Experienced engineers are expensive, venture funding is becoming more selective and large enterprises are increasingly capable of building their own software. Some startups will discover that the multinationals they hoped to sell to have become their strongest competition.

Others may benefit from precisely the same development. A startup can now find sophisticated enterprise customers, technical expertise and opportunities for commercial testing within India that previously required relationships in the US or Europe.

Over time, employees trained inside GCCs may establish companies of their own, bringing years of experience in specialised technologies and global markets. Some of these businesses may eventually sell products back to the corporations where their founders once worked.

The question for India is what proportion of this expanding engineering capability can be converted into independent companies with valuable products, customers and intellectual property of their own.

Creating millions of technology jobs is an important achievement, and GCCs deserve credit for their contribution. But employment alone cannot be the final measure of success for a country that also wants to become a global centre for technology entrepreneurship.

The next phase of India's startup story may depend less on how many engineers the country produces and more on what those engineers are able to build, who pays for their work and who ultimately owns the businesses created from it.

India has become an important place for the world's largest companies to develop technology. The harder task is ensuring that it also remains an attractive place to start and grow technology companies that belong to the people building them.