3 min read Last Updated : Jun 13 2025 | 11:04 PM IST
Pure-play generative AI (GenAI) projects, which are often business transformation programmes and small in deal sizes, are not seeing cancellations or pullbacks despite an uncertain macroeconomic environment, Infosys Chief Technology Officer (CTO) Mohammed Rafee Tarafdar has said.
Many companies had flagged client or sector slowdown during the fourth quarter earnings and analysts were worried about a sluggish discretionary spending environment due to fear of tariffs. However, Tarafdar dismissed all these concerns.
“We see clients continuing to invest in AI. A lot of the pure-play AI projects that have started with clear value propositions are moving forward. So, at this point, at least I have not seen any change in that,” he said in an interaction with Business Standard.
There are two types of AI and GenAI deals. One deals with existing projects which are currently re-designed with Gen AI to bring in more productivity and lower cost. The other is small GenAI deals which have, for most players, been in the proof of concept (PoC) stage with a few moving to industry deployment.
Tarafdar, however, said the time for such PoCs is over as GenAI is fairly scaled and mainstream now and agentic AI is built on GenAI capabilities.
“There’s no point in doing a PoC because we know what works, what doesn’t. So, it’s all about scaling. And today, I don’t think any of our customers are even questioning that. It’s largely about driving better adoption and better value from the GenAI implementations. That’s where the focus has now shifted,” said Tarafdar.
That will come as a relief to Indian IT service providers. Tata Consultancy Services (TCS) CTO Harrick Vin had told Business Standard earlier this year that less than 15- 25 per cent of all use cases are going into production. “The uptick is really slow, almost a trickle.”
Infosys, in its latest annual report, said that it delivered over 400 GenAI projects by bringing Infosys Topaz, its generative and agentic AI-powered services and solutions. “Our AI work spans a wide spectrum of priority areas like process improvement, engineering, customer service, cybersecurity, and employee productivity. We have built four small language models for banking, IT operations, cyber and enterprises broadly,” said the report.
Tarafdar added that the company is deploying AI agents at various levels. This includes task-level automation.
“It’s also happening on transform, which is where we are re-engineering the existing business process, which means rethinking in an AI native manner.”