Software quality from India.

Where the reputation came from

The stereotype formed during the body-shopping era, when the dominant model was supplying large numbers of relatively junior engineers billed by headcount against detailed specifications written elsewhere.

That model produced predictable outcomes. Engineers implemented specifications without context, had no authority to question requirements, and rotated frequently. The resulting software was often technically compliant and practically poor.

The reputation was earned. It was a reflection of a business model rather than of engineering capability, but from the client's side that distinction was invisible.

What has genuinely changed

Two things changed the picture substantially.

Global product companies built large engineering centres in India, and they hired for the same bar they used elsewhere. A generation of engineers now has experience of owning production systems rather than implementing specifications.

Remote work normalised distributed teams. An engineer in Bangalore can work directly with a product manager in Berlin without an intermediary layer translating requirements, which removes the structural cause of much of the old quality problem.

The variance remains wide. The ceiling is now high, and the floor has not moved much, which is why evaluation matters more than location.

What to check, regardless of geography

• Can you interview the specific engineer who will do the work?
• Will they push back on a requirement that does not make sense?
• Do they write tests without being asked, and can they explain what they chose not to test?
• Can they explain a past technical decision they now regret?
• Do they ask about your users, or only about your tickets?
• Is the engineer dedicated to you, or shared across clients?

The question that reveals most

Ask a candidate to describe something they built that did not work well, and what they would do differently.

Engineers who have genuinely owned production systems answer this immediately and specifically, because they remember the incident. Engineers who have only implemented specifications tend to answer vaguely, or describe a problem that was somebody else's fault.

This single question sorts more effectively than any technical quiz, and it works identically regardless of where the engineer is located.

The honest summary

India has excellent engineers and poor engineering vendors in roughly the proportion you would find anywhere, with a wider spread than most markets because the sector is so large.

The variable that predicts quality is not the country. It is whether the arrangement lets a senior engineer take ownership, ask questions and be accountable for the outcome. Evaluate for that, and the geography becomes a question about cost and timezone rather than about quality.