"... The external environment is hostile, more than what they had estimated and change has been more rapid," former director and chief financial officer of IT major Infosys, T V Mohandas Pai, told PTI when asked about repeated downward revision of guidance by leading IT players due to fall in export revenue growth.
This is a trend, say analysts, will continue.
On the impact of President-elect Donald Trump, Brexit and possible trade barriers on Indian IT, the technology investor said he expects it to be "very marginal".
On whether he saw Indian IT cannibalising its services, the chairman of Manipal Global Education Services and Aarin Capital Partners said: "Services have been cannibalised for long, it is accelerating now. What Indian IT companies need to do is to focus on future, identify trends, invest in advance and lead change rather than react to change."
Also Read
Stressing that Indian IT is already "rebooting" itself, Pai said the companies in IT and software services would "sort out" the stress they are facing as they are a globally dominant industry in a very competitive space under pressure.
Internet of Things, cloud computing and the overall move towards going digital, he said: "To catch trends as they emerge and invest ahead of need, to lead than be led, to innovate faster to lead the market."
Pai does not see any significant increase in IT manpower working on back-end applications, but also not an overall reduction of installed base as new work needs to be done. "As more back-end work gets automated, more work also becomes the back end. (But) growth of manpower is coming down," he added.
On doomsday prophecy of "death of code", Pai said: "Code will continue, huge legacy of around USD 4 trillion of code to work on. A large part of new code is getting automated, but building blocks (are) available. The breadth of need is so large that code will never die but can be done easier because of building blocks and automation."
"Training a computer by AI, ML (artificial intelligence, machine-learning) needs large number of cases to be out into databases to enable pattern recognition. Analytics would be needed on top and interpretation of analytics and decision making. Rule-based work would be diminished," he added.