TCS, Infosys reveal big AI plans in Q4 FY23, expect revenue streams to open up

Indian IT services giants Tata Consultancy Services (TCS) and Infosys have revealed major plans to monetise generative AI capabilities in their Q4 FY23 earnings call last week. This comes at the back of growing popularity of OpenAI’s chatbot, ChatGPT.

On Thursday, during the Q4 FY23 earnings call, Infosys CEO and MD Salil Parekh said, “We anticipate generative AI to provide more opportunities for work with our clients, and to enable us to improve our productivity.”

Similar thoughts were shared by N Ganapathy Subramaniam, COO of TCS. He said, “I think generative AI as a technology is very interesting. We have been working on it for some time, and we have done some pilots internally.”

Infosys’ Parekh also mentioned that the company is currently working on several projects which are using their AI capabilities.

“We are leveraging generative AI capabilities for our clients and within the company. We have active projects with clients working with generative AI platforms to address specific areas within their business,” Parekh said.

It is worth noting that Infosys, along with other investors, contributed $1 billion to OpenAI in 2015. OpenAI trained its chatbot ChatGPT on the GPT language model, the latest version of which is GPT4.

Parekh explained that their AI models are being trained on their internal software libraries.

“We have trained open source generative AI platforms on our internal software development libraries,” Parekh noted.

But can the AI tools created by Indian IT services companies take on the likes of ChatGPT, NotionAI, Midjourney, etc? Most likely not, experts, as well as the company management, say.

For starters, experts believe Indian IT might be a bit late to the AI party, and might be better-suited building on AI systems instead of building it themselves from scratch,

Kotak Institutional Equities noted in one of its reports, "With foundation models that reduce the cost and effort in building AI systems from scratch, enterprises can focus more on building with AI instead of building AI."

The company management highlighted that the tools that they would make would most likely cater to the needs of their clients, and would have specific expertise areas.

Subramaniam said, “I think we will build a certain expertise around AI capability around the tool set and stay relevant to our clients’ expectations.”

In a conversation with Business Today last week, the TCS COO discussed various automation tools the company has created around their clients needs.

He said, “TCS BaNCS, 95 per cent of the code of TCS BaNCS is completely generated based on a very high level of prompt that we do in terms of the higher level specifications that we write.”

Also Read: TCS Q4 FY23: Why did the company's QoQ sequential growth miss the mark? - BusinessToday

Also Read: TCS, Infosys Wipro: How Credit Suisse, SVB collapse can impact Indian IT sector - BusinessToday