From Hosting to Shaping: Building India’s Compute Economy Through Multi-Layer Coordination

Uber CEO Dara Khosrowshahi's five-day visit to India this week was a landmark in India's technology economy. The visit involved government engagement, strategic investment, tech capacity expansion, and showcasing of affordable mobility options amid competition. The two most important takeaways of his visit, to me, were:- 

New GCCs: Plans to open two major technology/engineering centres in Bengaluru and Hyderabad by end of 2027, accommodating ~9,600 new employees to support operations and product development. Hyderabad already hosts Uber’s first engineering facility outside the US. 

First data center in India: Partnership with Adani Group (AdaniConneX) for a data center, expected to be online in Q4 2026, enabling scaled tech testing and global deployment. Announced after Dara Khosrowshahi met Gautam Adani.

At first glance, these may appear to be separate developments — one involving office expansion, the other involving cloud expansion.

They are not.

Together, they reveal something much larger: artificial intelligence is beginning to reorganise the relationship between compute infrastructure, enterprise operations, urban development, engineering talent, and state capacity.

India’s policy architecture, however, still treats these as separate domains.

GCC policy operates separately from data-centre policy. Startup policy operates separately from engineering education. Urban development operates separately from AI infrastructure. Industrial policy operates separately from compute infrastructure.

This fragmentation increasingly reflects yesterday’s economy rather than the one now emerging.

The AI era is not merely producing new software products. It is creating a new infrastructure layer around which economic systems themselves may reorganise.

The strategic question for India, therefore, is no longer whether it will participate in the global compute economy. It will.

The more consequential question is whether India develops the institutional and infrastructural architecture required to shape parts of that ecosystem — rather than merely hosting externally controlled systems.

GCCs Are No Longer Back Offices

For nearly two decades, Global Capability Centres in India were primarily understood as back-office and operational support systems for multinational corporations.

That architecture is now changing.

Increasingly, GCCs are evolving into:

- AI engineering hubs

- cloud and platform operations centres

- infrastructure management teams

- cybersecurity ecosystems

- enterprise integration and deployment centres

This shift is especially visible among large technology companies.

Uber’s recent India announcements illustrate this transformation clearly. The company is not merely expanding office capacity. It is deepening operational capability in India while simultaneously partnering on data-centre infrastructure.

In other words, enterprise expansion and compute infrastructure are beginning to converge.

This convergence matters because AI-intensive GCCs have fundamentally different operational requirements from older-generation IT-service ecosystems.

They increasingly require:

- high-performance compute access

- cloud and data infrastructure

- low-latency systems

- secure digital environments

- AI deployment capability

- long-duration operational stability

However, not all companies possess identical capacities to build such ecosystems.

The Emerging Compute Divide

A bifurcation is beginning to emerge within the global technology and enterprise ecosystem.

On one side are hyperscale global companies such as Google, Microsoft, Amazon, and Uber. These companies possess:

- massive compute requirements

- long-duration India expansion plans

- substantial capital resources

- vertically integrated operational models

Such companies are increasingly likely to:

- build

- co-build

- or exclusively lease-in 

large hyperscale data-centre clusters in India.

This trend is already becoming visible.

For these companies, India is no longer merely a labour market. It is becoming a strategic operational geography for AI deployment, infrastructure expansion, and enterprise integration.

But this hyperscale layer represents only one segment of the emerging compute economy.

A second category of companies is equally important.

This includes:

- global startups expanding into India

- mid-sized multinational enterprises

- specialised technology companies

- and large Indian enterprises seeking AI integration

These entities would increasely require:

- technology/engineering capacity expansion 

- reliable compute/cloud capacity

- managed office/GCC spaces 

- enterprise-grade digital systems

- AI deployment environments

Yet many of them may not possess either the scale or the economic logic required to build exclusive hyperscale infrastructure independently.

They will instead require:

- ready-to-rent compute environments

- modular data-centre capacity

- professionally managed office ecosystems

- interoperable enterprise infrastructure

This creates a major strategic opportunity for India.

Data-Centres as Dynamic Institutions

India’s rapidly expanding data-centre sector should no longer be viewed merely through a real-estate or storage lens.

Data-centres are increasingly becoming dynamic institutions.

They influence:

- enterprise-location decisions

- AI deployment capability

- cloud operations

- startup ecosystems

- industrial AI systems

- cybersecurity architecture

- urban development patterns

- and regional economic geography

Simultaneously, a new category of Indian infrastructure actors is emerging.

Large real-estate and infrastructure companies are increasingly positioning themselves as gigawatt-scale compute infrastructure builders.

Their role extends far beyond constructing physical facilities.

They are becoming aggregators of:

- electricity supply 

- transmission systems

- cooling infrastructure

- land acquisition

- water systems

- networking infrastructure 

- and long-duration operational ecosystems

This aggregation capacity is strategically important.

Many smaller global companies, startups, and Indian enterprises cannot individually negotiate and manage these upstream requirements at scale.

Large Indian compute infrastructure builders, however, can.

This is where the integration amongst:

- GCCs

- data-centres

- office ecosystems

- enterprise systems

- and AI deployment

becomes economically and strategically significant.

The objective is no longer merely to build isolated data-centres or isolated office parks.

It is to build interoperable compute ecosystems.

Compute-Led Urbanisation

The implications of this transition extend beyond technology infrastructure.

Compute infrastructure increasingly behaves as anchor infrastructure.

Around compute clusters emerge:

- office infrastructure 

- housing infrastructure 

- logistics hubs 

- startups

- service ecosystems

- industry-academia partnerships

- and eventually physical industrial ecosystems

This represents a different organising principle for urban and economic development.

It can be understood as compute-led urbanisation.

Historically:

coal shaped industrial regions

ports shaped commercial cities

highways shaped suburban expansion

oil shaped twentieth-century geopolitics

In the emerging AI era, compute infrastructure may begin shaping economic geography itself.

This shift is unlikely to remain confined to metropolitan office districts.

At present, much of India’s compute demand is linked to:

- GCCs

- cloud platforms 

- digital services

- and office-based enterprise operations

Over time, however, AI’s deeper economic effects are likely to emerge within:

- mining

- refining 

- manufacturing

- warehousing

- transportation 

- energy systems

- and industrial corridors

As AI increasingly enters physical production systems, industrial corridors themselves may gradually evolve into compute corridors.

This transition has major implications for India’s long-term development strategy.

India’s Real Challenge: Coordination

India’s challenge in the AI era is not simply a shortage of talent, capital, or infrastructure.

It is a coordination problem.

India already possesses:

- engineering talent at scale

- growing startup ecosystems

- expanding renewable-energy capacity

- large infrastructure companies

- global enterprise interest

- engineering institutions

- and a massive domestic market

However, these systems often operate in fragmented silos.

GCC policy is separated from data-centre policy.

Engineering education remains weakly integrated with live enterprise ecosystems.

State startup missions frequently operate independently of compute infrastructure planning.

Urban development and industrial strategy are rarely linked to AI infrastructure.

The AI era increasingly rewards ecosystem coordination rather than isolated sectoral expansion.

This is especially important because markets alone may not solve the problem efficiently.

Large global companies can internally coordinate infrastructure, operations, talent pipelines, and partnerships.

Smaller companies and startups often cannot.

Similarly, individual engineering colleges cannot independently build deep partnerships with fragmented startup ecosystems.

Without coordination architecture, the system risks evolving into:

- isolated hyperscale enclaves

- fragmented startup ecosystems

- shallow placement-driven educational systems

- and uneven regional development

India therefore requires a multi-layer coordination framework.

The Central Government as Strategic Architect

The Union government must increasingly operate not merely as a regulator or incentive provider, but as a strategic architect of India’s compute economy.

This role would involve coordinating amongst:

- global companies seeking expansion in India

- Indian gigawatt-scale data-centre builders

- office-space developers

- elite central engineering institutions

- and interested state governments

The objective is not simply attracting investment.

It is architecting interoperable compute ecosystems.

This distinction matters.

The Centre’s role is not only to facilitate hyperscale infrastructure for companies capable of building their own ecosystems.

It must also help architect shared compute environments for startups and companies unable to independently build hyperscale infrastructure.

This includes:

- ready-to-rent compute capacity

- integrated office ecosystems

- interoperable enterprise infrastructure

- startup integration pathways

- and engineering talent ecosystems

In this framework, engineering institutions also become ecosystem participants rather than detached graduate suppliers.

Elite engineering institutions such as IITs, NITs, IIITs etc can become integrated into GCC and compute ecosystems through:

- infrastructure engineering collaboration 

- cloud and cybersecurity collaboration

- applied AI partnerships

- and real-world deployment simulations 

The Central government's role, ultimately, is to ensure that compute expansion translates into domestic capability accumulation rather than shallow hosting.

State Governments as Regional Ecosystem Architects

If the Centre operates at the national strategic layer, state governments operate at the regional and operational layer.

This distinction is important because economic ecosystems ultimately become local.

States possess direct influence over:

- land

- electricity

- transmission systems

- water access

- local startup ecosystems

- engineering colleges

- industrial corridors

- and urban-development systems

Therefore, state governments are best positioned to function as regional ecosystem architects.

In this framework, states coordinate amongst:

- local GCC leaders 

- local data-centre leaders 

- local office-space providers

- local startups and SMEs

- state engineering institutions

- and land and water stakeholders

This would create, what I call, compute federalism.

Different states can evolve different compute-development specialisations based on:

- industrial structure

- energy profile

- educational depth

- logistics networks

- and enterprise ecosystems

This diversity can become a strategic advantage, provided the Central government keeps a bird's-eye view and manages accordingly. 

Importantly, states can also ensure that compute infrastructure does not become an enclave economy disconnected from local enterprises and institutions.

States can negotiate:

- reserved compute capacity for state startups and local enterprises

- local supplier integration

- AI labs within state engineering institutions

- and local industry-academia partnership frameworks

This is critical because state capability formation increasingly depends on interaction with live ecosystems. Tier 2-3 engineering colleges can gradually evolve from mere talent suppliers to innovation partners, while local startups and enterprises can become more productive with assured compute and work spaces. 

Conclusion: From Hosting to Shaping

Many countries will likely host AI infrastructure in the coming decades.

Far fewer will shape:

- enterprise ecosystems

- capability formation

- institutional depth

- distributed innovation

- and long-duration technological leverage

India’s strategic opportunity lies not merely in attracting data-centres or GCCs individually.

It lies in embedding compute expansion into broader capability formation.

That requires:

- institutional coordination

- infrastructure integration

- state-level ecosystem building

- educational linkage

- enterprise diffusion

- and long-term operational depth

The larger opportunity is not simply economic. It is paradigmatic.

Compute is no longer merely technological infrastructure.

It is increasingly paradigmatic infrastructure capable of reorganising:

- urbanisation

- industrial geography

- enterprise systems

- startup ecosystems

- educational institutions

- and state capacity itself.

The industrialisation of AI is no longer a future abstraction.

It is already beginning to shape capital allocation, infrastructure planning, enterprise expansion, and institutional strategy in real time.

India will certainly participate in this transition.

The more consequential question is whether it develops the coordination architecture required to shape the ecosystem around it — or merely becomes a hosting geography for systems designed elsewhere.