Bhubaneswar: Can It Become an AI City?

Bhubaneswar is already a tech city. That is the starting point, not the conclusion. The Software Technology Parks of India node has been operational here since 1991 — earlier than most Indian cities that now carry more prominent technology reputations. Infocity, the dedicated IT park, has been drawing mid-size technology operations for two decades. KIIT University produces more than 30,000 engineering students annually. Wipro, Infosys, TCS, and Zensar have established presences. The cost structure sits well below Bengaluru and Hyderabad equivalents.

A city that has been building technology infrastructure for thirty years is not starting from zero. The question is whether the next phase — AI-native companies, machine learning infrastructure, applied research institutions, the specific ecosystem components that would make an AI company choose Bhubaneswar over an alternative — is achievable. And what it would take.

What a tier-2 tech city actually means

The phrase "tier-2 tech city" is used in India to describe cities that have technology ecosystems but not the gravitational pull of Bengaluru or Hyderabad or Pune. What it actually describes is a city that has the inputs — talent, infrastructure, cost structure — without the compounding effects that come from ecosystem density.

Bengaluru is not a better city to build a technology company in because it has better infrastructure than Bhubaneswar. It is better because it has more companies, which means more engineers who have worked at those companies, which means more people who know how to build what you're trying to build, which means more capital from people who have made money from those companies, which means more ambition from people who have seen what ambition can produce.

This is a compounding dynamic. It took Bengaluru decades to reach its current state, and it took triggers — the early Infosys and Wipro decisions, the specific government policies of the 1990s, the physical concentration of talent in a single geography — that were partly deliberate and partly fortuitous.

Bhubaneswar is behind that curve. The question is whether it is on the curve at all.

The talent case

The talent pipeline is real and underrated.

KIIT and IIIT Bhubaneswar produce a significant volume of engineering graduates annually. IIT Bhubaneswar, still establishing itself but with the institutional credibility of the IIT network, adds a different tier of output. The students coming out of these institutions are, in aggregate, competitive with students from equivalent institutions anywhere in India.

What is less developed is what happens to them. A significant proportion leave — to Bengaluru, Hyderabad, Pune — not because they would prefer to but because the density of opportunity in those cities makes staying in Bhubaneswar feel like a step down. The city produces talent that other cities convert into ecosystem.

This is the core problem with tier-2 tech cities, stated plainly: they invest in talent production and export the output to places with more developed ecosystems. The talent is the asset; the city is not capturing the compounding.

An AI city requires a different dynamic. It requires anchor companies that keep senior engineers in the city. It requires career paths that don't require leaving. It requires the specific signal — visible to a 22-year-old making a decision about where to build a career — that staying here is not a compromise.

The infrastructure that already exists

STPI Bhubaneswar established early and has grown. The Infocity corridor has functional tech park infrastructure. The Smart Cities Mission investment has improved urban infrastructure in ways that matter to knowledge workers: connectivity, transit, urban livability. AIIMS Bhubaneswar has upgraded healthcare access. The airport has expanded.

What the city has built is the infrastructure substrate — the conditions under which a tech ecosystem can function. What it has not yet built is the institutional density that makes the ecosystem self-reinforcing.

A specific gap: AI infrastructure. Data centres, GPU compute capacity, the specific physical infrastructure that AI training and inference requires, are concentrated in a small number of Indian cities. Bhubaneswar is not yet one of them. This is not a permanent condition — it is a capital allocation decision that has not yet been made. But it matters for the AI city question specifically. AI companies require compute infrastructure that is proximate, affordable, and reliable. The absence of this in Bhubaneswar creates a friction that has to be overcome by other advantages.

What "AI city" actually requires

The phrase gets used in India to describe cities that are investing in AI policy, AI education programs, and AI-themed industrial parks. Most of what is described under this label is not an AI city — it is a city with AI branding.

An actual AI city has specific characteristics. It has a concentration of AI-native companies — companies whose core product is built on machine learning, not companies that have added AI features to existing products. It has applied research institutions that are producing work that those companies can use. It has the career infrastructure — senior AI roles, competitive compensation, visible paths to seniority — that retains AI-focused engineers rather than exporting them. And it has the capital infrastructure — investors who understand AI businesses, who have made and lost money in the sector, who can evaluate what is being built with real sophistication.

By this definition, there are very few genuine AI cities in India. Bengaluru is closest. Hyderabad is building one. The others are varying distances behind.

Bhubaneswar's path to this list is not through competing directly with Bengaluru on the same terms. It is through a specific bet: that the next phase of AI development in India will distribute across more cities than the current phase, and that a city with existing talent infrastructure, a cost advantage, and deliberate policy investment can be a destination for that distribution.

This bet is not irrational. The history of technology ecosystems in India is a history of distribution — from a handful of cities in the 1990s to a much larger set today. The cities that benefited from each phase of that distribution did so because the conditions were in place when the capital started looking for new locations.

The honest answer

Bhubaneswar can become a serious technology city. It is already one in partial form. Whether it becomes an AI city specifically depends on decisions that haven't been made yet — about compute infrastructure, about anchor AI companies, about the specific policy interventions that would make it a preferred location rather than a default alternative.

The talent is there. The cost structure is there. The foundational infrastructure is there.

What is not yet there is the trigger — the specific set of decisions that tips a city from a place where technology companies can function into a place they choose first. That trigger is not inevitable. It requires active investment in the specific things that are missing, not passive confidence that what exists is sufficient.

The window is open. It will not stay open indefinitely. The cities that become AI destinations in this decade will be the ones where that investment happens while the cost structure still makes the bet worthwhile.

Bhubaneswar is on the list of places where that bet makes sense. Whether it makes the list of places where the bet gets made is still to be determined.