The infrastructure question every AI strategy must answer

Artificial intelligence (AI) may be dominating boardroom discussions, but companies are discovering a less obvious constraint: Their infrastructure is not ready for it.

While 77 per cent of global enterprise leaders see AI as a board-level priority, only 29 per cent say their infrastructure can support it at scale, according to a report by Tata Communications and Bloomberg Media Studios.

For Mr Amitabh Sarkar, Tata Communications’ vice-president and head of Asia Pacific and Japan (Enterprise), the findings point to a widening gap between enterprise ambition and readiness. “Infrastructure is no longer a backend concern. It’s now a CEO agenda,” he said.

The challenge, he argued, is not a lack of AI ambition, but the complexity of the environments in which AI must operate.

Over the years, organisations have added cloud platforms, software applications, cybersecurity tools and data systems to solve business needs. While each investment serves a purpose, the result is often a fragmented technology landscape that is difficult to manage at scale. Without clear visibility of how data moves across users, applications, data centres and cloud environments, organisations may struggle to govern AI systems, manage risk and build trust.

The complexity also makes it harder to move projects beyond the pilot stage. “Whether you’re provisioning infrastructure, integrating legacy systems or applying security policies, these all take too long,” said Mr Sarkar. “That creates a gap between AI experimentation and AI industrialisation.”

KEEPING CRITICAL OPERATIONS RUNNING

For Tata Communications, addressing AI readiness starts with what it calls the Digital Fabric, a unified layer that connects networks, cloud environments, applications, security systems and users. That foundation becomes especially important in industries where even brief disruptions can have real-world consequences.

Take, for instance, an airline. A connectivity issue can affect everything from passenger check-ins to baggage handling, creating knock-on effects across the travel journey.

“If connectivity becomes an issue, the platform’s multi-path routing automatically takes an alternate path, whether secondary or tertiary, routing it to our own ecosystem of partners,” explained Mr Sarkar. “This ensures the transition is seamless and transparent to the end customer, with no downtime or interruption for the airline and its passengers.”

To support such mission-critical environments, Tata Communications launched IZO Data Centre Dynamic Connectivity, a software-defined platform that links major data centres across five continents. Using deterministic multi-path routing, it can deliver predictable latency and performance while rerouting traffic when disruptions occur. This enables enterprises to maintain more than 99.99 per cent service availability across mission-critical infrastructure, according to Tata Communications.

The platform also gives organisations greater visibility into network performance, allowing them to monitor traffic, receive proactive alerts and scale bandwidth as workloads evolve. Its self-healing capabilities can identify faults, isolate affected segments and restore services with minimal manual intervention.

The goal, said Mr Sarkar, is not simply to recover faster when disruptions occur, but to prevent them from affecting operations in the first place.

Resilient network design can help organisations remain operational during unexpected events. “We’ve seen multiple unpredictable events happening across the globe, but our customers remained up and running,” he said. “Our always-on approach and resilience by design are fundamental to our Digital Fabric, whether across networks, security or cloud environments.”

WHERE DOWNTIME IS NOT AN OPTION

Tata Communications is extending its Digital Fabric through strategic partnerships designed to solve operational challenges in sectors where resilience is critical.

One example is its collaboration with Cisco. By integrating Tata Communications’ global eSIM orchestration solution Move with Cisco’s IoT Control Centre, enterprises can manage connected devices across SIM providers, private and public networks and eSIM standards through a single platform.

The result is a more unified approach to managing large-scale Internet of Things (IoT) deployments, particularly in sectors such as transportation, logistics and manufacturing, he noted. “This reduces integration effort and helps organisations avoid lock-in,” said Mr Sarkar. “That translates into greater visibility, faster response times and better operational outcomes, especially for organisations managing large fleets of connected devices.”

The same principle underpins Tata Communications’ partnership with Johnson Controls. Its Digital Fabric provides the infrastructure supporting the OpenBlue building intelligence platform, enabling predictive AI, secure connectivity, IoT telemetry, edge computing and cloud orchestration. The partnership has helped deliver energy savings of up to 12 per cent across sites in Singapore and 10 per cent in London through autonomous building management.

For Mr Sarkar, these examples point to a broader lesson for organisations looking to scale AI. One of the biggest mistakes organisations make, he said, is treating AI as a technology deployment rather than an enterprise architecture challenge.

“Instead of asking which AI tool to use, the better question is: Which business processes are we trying to transform?”, said Mr Sarkar. “The issue is not whether enterprises want AI – they clearly do. The issue is whether the infrastructure can carry the AI at scale.”

Learn more about partnering with Tata Communications to build AI-ready digital infrastructure.