'I look to China as gurus': Zoho's Sridhar Vembu

Vembu’s central lesson is that India should build and retain ownership of important technologies. Engineering talent alone does not guarantee lasting prosperity if foreign companies own the products and intellectual property. He believes China offers stronger lessons than America because it developed domestic platforms and businesses despite earlier economic constraints.

He pointed to Baidu and WeChat as examples. China supported platforms that initially lagged international competitors, then used them to build local capabilities, large user bases and valuable data ecosystems. Those foundations supported later innovation, including TikTok and advances in AI.

This argument matters as AI changes software work. Vembu said India’s services model could become more vulnerable if automation reduces the number of engineers needed. He said Zoho is redeploying engineers toward new technology capabilities, while India should solve its own problems and develop products it controls.

Technology ownership means controlling the intellectual property behind a product, such as its software, designs or platform, and capturing the economic value it creates. The owner decides how the technology is developed, commercialised and improved. This can create lasting capabilities and profits within the country.

Providing software services is different. Indian engineers may design, code or verify technology for companies such as NVIDIA, Microsoft, Oracle or Cisco, while those companies retain ownership of the resulting products and intellectual property. India receives service revenue, but may not capture a proportionate share of later profits.

Vembu’s point is not that services lack value. It is that services alone may leave India dependent on foreign-owned products and platforms. He argued that building domestic technology ownership would let Indian companies accumulate capabilities, users, data and profits, creating a stronger foundation for future businesses.

Vembu claimed that approximately 40–50% of NVIDIA’s chip design work happens in India. He included substantial verification work in that estimate. He also pointed to significant software development in India for Microsoft, Oracle and Cisco. These claims illustrate his broader argument that India performs major technical work without capturing an equivalent share of resulting economic value.

The key mechanism is separation between engineering labour and ownership. Indian teams may contribute heavily to designing, testing or developing technology, while the foreign company owns the intellectual property, product and profits. Vembu suggested that if India captured half the value of such work, Karnataka’s GDP could be two or three times its current size.

The figures require caution. The article states that Vembu’s estimates were not independently verified. They are therefore claims made during his keynote, not established measurements. Even so, they support his call for India to build and own more technology rather than mainly provide services.

Chinese platforms helped create a domestic technology base that China could control and improve over time. Vembu cited Baidu as an alternative to Google and WeChat as a central part of China’s digital ecosystem. Their importance came from building products for local users instead of relying only on foreign platforms.

The mechanism worked through three connected assets. Domestic platforms developed technical capabilities, attracted large user bases and generated data ecosystems. Those assets gave companies experience with products, infrastructure and user needs. Vembu said this foundation was necessary for the next stage of innovation, even though the platforms initially lagged international competitors.

The effects extended beyond the original services. Vembu linked the foundation to globally competitive companies such as TikTok and to advances in AI. His lesson for India is to invest in underlying technologies, not focus exclusively on the newest AI models. Building that base requires years of investment and patience.

If AI writes much of the code now produced by engineers, India’s traditional software-services industry could face major disruption. Vembu said AI already writes better code than 60–70% of software engineers and could outperform as many as 90% within three to six months. These are his projections, not independently established benchmarks.

The mechanism is labour substitution. If fewer engineers are needed to build software, companies may buy less routine coding and software-development capacity from Indian service providers. Vembu compared the risk to the handloom industry before mechanisation, warning that automation could dramatically reduce the number of workers required.

He said Zoho was not responding by laying off employees. Instead, it was redeploying engineers to build other technology capabilities. Vembu’s wider implication is that India should use its engineering strength to own underlying technologies and solve its own problems before AI reduces the value of services work.

Owning a technology product can generate more long-term value because the owner controls the intellectual property and the business built around it. The owner can keep improving the product, decide how it is used and capture value from its growth. By contrast, building technology owned by foreign companies mainly produces service income for the country doing the work.

Vembu used India’s engineering contributions to global companies as the contrast. Indian teams may design chips or develop software, but NVIDIA, Microsoft, Oracle or Cisco retain the products and intellectual property. China took a different path by investing in domestic platforms such as Baidu and WeChat, even when they initially lagged global alternatives.

Those platforms created capabilities, user bases and data ecosystems that supported later innovation. Vembu linked that foundation to companies such as TikTok and advances in AI. He argued that India needs similar long-term investment, even if domestic products take years to become competitive.

Intellectual property gives a company control over the technology it owns. A digital platform then provides a place to deploy that technology and attract users. As more people use it, the platform can develop stronger capabilities and a larger data ecosystem. Together, these assets help a country retain more value and build expertise.

Vembu’s examples were Baidu and WeChat. Baidu became a Chinese alternative to Google, while WeChat became central to China’s digital ecosystem. Their platforms created technology capabilities, user bases and data. That foundation supported the next stage of innovation, including globally competitive companies such as TikTok and advances in AI.

The lesson for India is to build underlying products and platforms, not only perform engineering work for foreign owners. Vembu argued that this requires years of investment and a willingness to support products that initially lag established alternatives. The payoff is greater control over future innovation and economic value.