Indian IT’s FDE race: Can elite tech squad scale? | India Business News

Indian IT firms are racing to build thousands of FDEs. The question is: can they do it at the pace required, because their older upskilling/reskilling models may not work here

Bengaluru: A forward-deployed engineer may be the most coveted job today. Every tech company looks to be hiring FDEs — highly skilled engineers who work directly with customers to identify problems, co-develop solutions and deploy AI at speed.Indian IT firms are racing to build thousands of FDEs. The question is: can they do it at the pace required, because their older upskilling/reskilling models may not work here.The FDE model, pioneered by US software company Palantir, puts engineers close to customers to identify problems, build solutions and own outcomes, rather than simply execute predefined specifications. Industry executives say it cannot be replicated simply by increasing onsite headcount or relabelling existing service roles.Infosys, for instance, plans to scale its frontier engineering team to 6,000 engineers, while TCS is reportedly building an 8,900-strong FDE pool. LTIMindtree has announced a programme to develop more than 1,000 AI-certified engineers, including FDEs.“Exceptional problem-solving does not scale linearly with brute-force headcount growth,” said Vishal Sikka, former Infosys CEO and founder of Hangten Systems. He said if IT services firms try to simply relabel onsite staff as FDEs while continuing to bill by the hour, it will miss what the new model really requires.“AI breaks the old link between revenue and headcount. This is India’s moment to move from supplying effort to amplifying human potential. We have talent. The question is whether we have imagination. I am convinced that we do have the imagination, in abundance,” Sikka said.Anjor Kanekar, a former Palantir FDE and founder of Platypus Technologies, said the challenge is that the role requires an unusual combination of engineering expertise and customer-facing judgement.“A conventional engineer builds a spec that someone else wrote. An FDE sits with the customer, works out what is worth building, and ships it into their environment,” he said. “The hard combination is production-grade engineering plus the judgement to walk into a messy business problem and decide what to build, plus the temperament to do that in front of a customer.”“Most strong engineers do not want to be customer-facing. Most customer-facing people cannot ship production code. Holding both, with the taste of knowing which problem matters, is the rare part,” Kanekar said.Viral Shah, CEO of JuliaHub, said Palantir’s organisational structure is another reason its model is difficult to replicate at the scale of Indian IT firms. “They usually operate on projects with much smaller teams than Indian IT companies, with very highly skilled young engineers well versed with modern tools and Palantir’s own products. They also have a lot of autonomy in decision-making around technologies, team structure, hiring, etc. which allows them to move fast.”Shah said Indian IT companies have adapted through successive technology shifts, but the AI transition is different because it changes the labour-arbitrage equation. “So long as the leadership teams in these companies are hands-on and deeply technical and themselves understand what AI-native teams look like, I have no doubt that they have the institutional capability to adapt to the changing world and what their customers need.”The talent challenge is compounded by a paradox in AI: while the technology is replacing significant amounts of traditional software work, there is simultaneously a shortage of AI-native talent. Shah said Palantir CEO Alex Karp has also spoken about the difficulty of hiring the right engineers.Cognizant chief learning officer Thirumala Arohi said the capabilities required for FDE roles cannot simply be created through conventional training. “The combination we need—technical depth, real domain ownership, and a particular way of thinking through problems—can’t be built on a training calendar,” he said. What matters is whether engineers can quickly understand an unfamiliar client domain and adapt as AI models and tools change.Cognizant has trained 350,000 associates in AI fluency, with more than 150,000 targeted through AI Bridge programmes. “Mass AI fluency raises the floor. It doesn’t, by itself, produce someone who can walk into a claims operation and own the outcome,” Arohi said. “Only a small number who come through that base clear all three gates, and that gap is intentional, not a shortfall.”The answer to scaling, therefore, may lie not in making FDE teams larger but in keeping them small and replicating the structure. Arohi said the model works because it removes handoffs between scoping, building and accountability.“We work in pods of one to three people… The pod size is a design principle, capped deliberately, because bigger teams bring the handoffs right back. When we scale, we add pods. We don’t grow the pod.”For Indian IT firms, Lalit Wadhwa, CTO of Coforge, said the test is whether engineers can discover and frame problems, build production solutions, support adoption and measure outcomes. “Replicating those capabilities requires engineering judgment, appropriate decision-making authority, and a structured feedback loop with platform teams.”Ramesh Lokre, co-founder and CEO, Saicon Consultants, said the challenge goes beyond hiring and training. “The challenge is not just hiring or training; it is ownership and enterprise mindset that make the role effective. If Indian IT firms treat forward deployment as the traditional onsite-offshore model under a new name, they will miss the point.”