By Sudhir Kunder
As AI transitions from early pilots to full-scale production systems, India is racing to shrink latency and bring compute closer to users — a prerequisite for real-time intelligence across industries. From optimising supply chains and customer experiences to autonomous workflows, the competitive edge now favours organisations that pair AI with robust, low-latency infrastructure close to where data is generated and used.
AI Adoption in India: Infrastructure Must Keep Pace
Global studies show broad experimentation with AI agents — and Indian enterprises are moving even faster. While surveys like McKinsey’s 2025 global AI adoption research highlight significant worldwide experimentation with agentic AI, insights from Indian consultancies suggest that over 80% of Indian organisations are exploring autonomous agents and multi-agent workflows, with nearly 70% reporting that GenAI initiatives meet or exceed ROI expectations. This underscores that the business case for AI in India is strong, but only where infrastructure keeps pace.
Latency Matters: Closer Compute = Better Outcomes
For mission-critical AI workloads from autonomous vehicles and industrial robotics to healthcare and safety systems, every millisecond in latency counts. When data, models, and applications are tightly coupled geographically, latency shrinks, real-time decisions accelerate, and energy efficiency improves. Indian deployments show that moving compute from distant cloud regions to nearby edge and local data centres can reduce end-to-end latency by 60–65%, a game-changer for reliability and AI quality.
Globally connected, locally constrained
What’s a nuisance in video streaming can become a hazard in real-time operations. More bandwidth or faster devices rarely fix it; the root cause lies in network topology – and in where infrastructure lives. Despite a globally connected Internet, geography still limits performance. The solution is to bring compute and interconnection closer to where businesses operate. That’s exactly what’s underway in India’s fast-scaling digital backbone: installed data-center capacity reached ~1.3 GW by mid-2025, with another ~2.9 GW expected by 2030; Mumbai leads as the Indian hub, while Chennai, Hyderabad, and Delhi-NCR are gaining ground to serve local, low-latency demand. Market trackers similarly forecast India’s total IT capacity to triple to >4 GW by 2030, with Mumbai and Chennai accounting for ~70% of today’s absorption – evidence that proximity is becoming a decisive location factor for AI workloads.
Internet Exchanges: Local data, local paths
Internet Exchanges (IXs) physically interconnect networks so SMEs, SMBs, Enterprises, Cloud Providers, Content Platforms & ISPs can exchange traffic directly, instead of detouring over the public Internet – keeping local data local and reducing latency. As AI pushes terrestrial infrastructure to its limits – and as businesses demand coverage “everywhere” – space-based connectivity becomes part of tomorrow’s digital ecosystem.
India is moving fast: Reliance Jio demonstrated Jio Space Fibre with SES to deliver gigabit satellite broadband to remote districts, illustrating how satellite backhaul complements fiber and 5G. Bharti-backed Eutelsat OneWeb secured Indian space-regulator clearance in late 2024 and has been seeking DoT approvals to activate its Indian gateways, aiming to provide low-latency LEO services across South Asia on a B2B basis. In 2025, Starlink received its unified license in India under strict data-sovereignty conditions – another signal that satellite broadband is entering the mainstream of national infrastructure.
For AI-enabled operations – from rural healthcare to maritime logistics – these developments make satellites a practical extension of terrestrial networks rather than a separate universe.
Building an interconnection layer for orbit
The remaining challenge is to bridge Earth and orbit with the same quality of interconnection we expect on the ground. DE-CIX and the German Aerospace Center (DLR) are working within the European Space Agency’s OFELIAS project to develop protocols, algorithms, and procedures that make optical (laser) satellite links more robust against clouds and atmospheric effects – so they can be integrated seamlessly into terrestrial networks. In parallel, DE-CIX’s Space-IX initiative outlines how satellite operators can interconnect with each other and with clouds and content platforms, enabling data and AI inference to flow across Earth-to-space infrastructure with minimal latency.
From Earth to orbit: Minimise latency, maximise efficiency.
The business case for AI is clear – cost savings, revenue growth, and faster innovation – but the precondition is an infrastructure stack that collapses the distance between users, data, and models. Global surveys confirm the agentic shift; India-specific studies show adoption is broadening as enterprises industrialise AI. At the same time, India’s interconnection and data-center capacity are scaling to host more data locally and shorten round-trips – key for low-latency AI services at the edge. Even payment networks and systems are embracing AI-driven risk scoring to keep transactions safe at a national scale, underscoring why high-quality, low-latency connectivity matters beyond convenience.
Latency First, Intelligence Everywhere
The Indian AI journey is accelerating from pilots to production across enterprises, startups, governments, and critical infrastructure services. But AI’s promise can only be realised when data, models, and compute are close to users and when networks are intelligently interconnected.
By scaling data centres, enriching Internet Exchanges, and embracing satellite connectivity as part of a larger digital ecosystem, India is building the low-latency, resilient infrastructure required for real-time AI at national and global scale.
The takeaway: To power intelligent industries on the ground and in the air in India, organisations must architect for proximity. Compute and interconnection don’t just accelerate performance; they unlock AI’s full value in real time.
Sudhir Kunder is the Chief Business Officer, DE-CIX India
