Forward Deployed Engineers are Testing Indian IT, Not Replacing It
By Ajay Rag ·
Every time a Silicon Valley giant announces another army of engineers it plans to embed inside client companies, Indian IT stocks fall.
OpenAI’s move in May wiped billions off TCS, Infosys, HCLTech, Wipro and Tech Mahindra in a single session. Microsoft’s July 2 announcement of a $2.5-billion ‘Frontier Company’ unit, staffed by 6,000 engineers, did much the same.
The market’s verdict has been consistent. The prevailing perception is that this marks the beginning of the end for the industry that built India’s over $300-billion technology export machine.
However, according to multiple industry executives AIM spoke to, this is a test of execution, not a death sentence. Whether Indian IT passes that test depends on decisions its leadership makes over the next 18 months.
“The rise of forward deployed engineers (FDEs) is fundamentally different, but not because services and implementation have suddenly become a new battleground,” says Gaurav Vasu, Founder and CEO of the research firm UnearthInsight.
What has actually changed, he notes, is that AI knowledge itself is becoming a commodity. Enterprises are no longer bottlenecked by model quality, but by the much harder problem of applying that intelligence inside real business processes and legacy systems.
A forward deployed engineer, in his framing, is the AI-era version of the traditional functional consultant. And on that turf, Indian IT still holds a significant advantage built over decades of client relationships and process knowledge—an edge AI-native companies cannot manufacture, only rent.
Spin the Wheel
In the latest quarterly earnings, TCS, India’s largest IT services firm, said it plans to increase the share of employees working under a forward-deployed-engineer-style operating model, although Chief Operating Officer Aarthi Subramanian said the company’s approach differs from the model popularised by AI-native software firms.
She described TCS’ ‘rapid-build engineer’ model as teams of multi-skilled specialists deployed in small squads to solve specific customer problems.
“At least 1% of our employee base, we would want to target to work in the new operating model,” she said.
Zoom In
The deployment teams of the big tech firms, according to Vasu, exist because AI adoption depends on integration and business outcomes—not because Redmond or San Francisco suddenly decided to compete with TCS.
For Microsoft, the goal is to accelerate adoption of its own AI platforms and cloud infrastructure. For AI-native labs like OpenAI and Anthropic, it’s about proving their models can survive contact with actual enterprise environments. Hyperscalers still need Indian IT’s delivery muscle at scale, just as Indian IT needs frontier models to build what clients now expect.
“The future is therefore likely to be defined by a combination of partnership and competition, rather than outright replacement,” says Vasu.
On whether hyperscaler-scale FDE hiring will threaten IT service firms Ed Kiesling, Chief AI Officer at Progress Software, says, “I think that this is a huge opportunity for services firms to move up the value stream and lead AI transformations at the clients they already serve.”
Kiesling adds that while some organisations are approaching AI purely as a cost-cutting exercise, a framing he calls “unfortunate”, the more common and more durable use case is enterprises using AI to expand what they can offer customers, not merely to shrink headcount.
Industry executives also note that the FDE wave is not a new phenomenon so much as an old one being repackaged.
Hyperscalers building professional-services arms, Praveen Bhadada, CEO and MD of Neovay Global, notes, is “standard practice at the turn of every disruptive technology wave”. The fact that AI labs feel compelled to do this is itself proof that scale ultimately requires systems integrators, not proof that integrators are being cut out.
Risk Factors
Where Bhadada does see genuine risk, it isn’t from Microsoft or OpenAI directly. It’s from what he calls platform lock-in.
Engineers with privileged access to a model provider’s roadmap and product teams will naturally pull enterprise clients deeper into that provider’s ecosystem, he argues. Ironically, that dynamic could eventually push enterprises back towards independent integrators that are better positioned to prevent single-vendor dependence.
He also points to a talent-acquisition crisis hiding in plain sight. It is “genuinely difficult”, he says, for services companies to hire true FDE-calibre professionals laterally. Indian firms will have to grow this capability internally by identifying the right people, creating deliberate learning paths, and deploying them in customer-facing, onsite roles where business problems actually exist.
Vasu points to traditional application development, testing and maintenance as the segments facing the sharpest disruption, since AI is automating exactly this kind of routine, hourly-billed work. Enterprises, he says, are shifting from paying for effort to paying for outcomes.
But he is equally clear that this doesn’t erase the value of domain expertise or enterprise integration. The opportunity, in his view, lies in Indian firms moving up the value chain into specialised AI applications and industry-specific solutions, with human oversight and enterprise architecture remaining essential even as lower-value coding work gets commoditised.
The Bigger Picture
Indian IT’s recent stock declines may be pricing in obsolescence when the more accurate story is closer to a deadline. Indian IT incumbents retain real, hard-to-replicate advantages in domain knowledge and enterprise trust, but those advantages have a visibly shrinking shelf life.
Vertical specialisation, internally developed FDE-calibre talent, outcome-based billing and deeper hyperscaler partnerships are, in this reading, no longer optional strategic ideas—they are close to the only path left to firms that still want to matter five years from now.