“If you were built simply on a cost arbitrage basis, and your primary access to value was, we are significantly lower cost than our onshore counterparts, you hire, and build a certain kind of talent. And in that context, everybody has now got that kind of talent,” Vaz said, adding there is a “full scale compete happening right now”.
The bigger issue, however, is the competition between people and machines, according to Vaz, who believes that the sector will face more pressures going forward given AI’s ability to automate tasks like code testing. “That competition now is also starting to happen, both within the GCC and within the IT services space. Now, in the short term, there will be pressure if you have not upskilled, reskilled, transformed in order to provide a very different kind of capability because that is what is going to be the expectation,” he said.
According to Vaz, while AI can do the work of thousands of people, there will be value in deploying them in a reskilled fashion for building the next generation of AI applications, though that transition is yet to happen and will also create pressures on cultures and margins.
Vaz, also an active investor in India’s startup ecosystem, said while Indian startups tended to replicate existing ideas and strive to compete on an execution basis, they are now starting to look at intrinsic problems in the Indian context and innovate to create solutions with AI.
“Unfortunately, a little bit comes from the legacy of what Indian IT has been about-it’s not been about where companies come to innovate, it’s been about where companies come to scale and to operate much better. I feel like that has to change and it’s starting to change,” Vaz said.While there is value in building diverse Indian large language models, it is expensive in terms of compute and storage, Vaz noted, adding that building on top of existing models, with small language models requiring less compute, to solve specific problems is where innovation will happen.Given the rising demand for compute, Vaz said the Indian government’s push towards manufacturing semiconductors is a very strategic shift from consumer to producer. While it is not something India has historically been known for, in the next five to ten years, when producing these chips could potentially become a national priority, it will be a helpful ability to have.
For Publicis Sapient, the company’s Indian GCC was conceived not as an offshore delivery centre but as part of a globally distributed delivery model, enabling access to a strong talent pool, Vaz said. It has the largest share of the company’s 23,000-strong global workforce.
