That acceleration is now reshaping everyday developer life.
Raghav Shetty (name changed on request), a .NET architect in Bengaluru with more than 17 years of experience, says engineers who understand how to use AI effectively will find their lives easier. Tasks that previously took two to three days can now be completed within a single day through better prompt engineering.
But that efficiency comes with caveats. AI agents may work correctly 90 per cent of the time, he says, but human intervention remains critical: Whether to fix errors, optimise solutions, or select the right output from multiple AI-generated options.
Mahajan adds that the shift has fundamentally altered the nature of coding itself. “Personally, we have stopped coding by at least 90 per cent. Most of our time now goes into prompts and reviews. Functions that were earlier handled by CTOs (chief technology officers) and managers are now becoming part of a developer’s role.”
These ground-level changes have fed into broader anxieties about the future of India’s IT services model. The latest trigger was Anthropic’s launch of Claude Cowork, a generative AI platform with 11 plugins introduced last month to automate tasks across legal, sales, marketing and data analysis. The move prompted investors to question how essential many legacy software layers would remain if AI agents could independently solve complex operational problems.
That concern was followed by a fall in stock prices and valuations across several large IT services players, pushing some commentators to write the obituary of the $225 billion industry.
“We acknowledge that it is difficult to quantify the real impact of AI (both deflationary and inflationary) and the duration of demand deflation given where we are in the cycle. However, it is overly simplistic to assume that AI can automatically generate enterprise-grade software and replace the value IT services firms create. Indeed, IT services companies remain the plumbers of the tech world, and if enterprise software or SaaS is rewritten on a bespoke basis by agents, it will still require significant services plumbing to function in an enterprise context and minimise AI slop,” said a JP Morgan note.
The industry has faced multiple crossroads over the past three decades: The Y2K challenge at the turn of the century; the aftershocks of the 2008 global financial crisis; repeated claims that IT companies would wither under the impact of digital technologies — social, mobile, analytics and cloud — and even the Covid pandemic.
“Every major inflection point in this industry — offshore, cloud, SaaS, automation — has triggered the same narrative: ‘This time services are dead’. Yet the sector adapts because enterprises do not buy technology, they buy outcomes. And outcomes require integration, orchestration, governance, change management and risk absorption,” said Phil Fersht, CEO and founder of HfS Research.
Analysts argue that the market’s mistake lies in confusing task automation with operating model transformation. Automating legal drafting, coding, testing or support workflows does not magically rewire data estates, security models, regulatory frameworks or the complex interdependencies of legacy systems across a vast enterprise.
However, companies will need to recognise that the traditional pyramid structure of the IT industry is becoming more fluid, and is likely to remain so as firms seek to do more with less. The emphasis is shifting towards increasing top-line growth with minimal additions to headcount in order to demonstrate AI-driven efficiency and productivity gains.
-
It is overly simplistic to assume that AI can replace the value IT services firms create -
Analysts argue that the market’s mistake lies in confusing task automation with operating model transformation -
Traditional pyramid structure of the IT industry is becoming more fluid
