Edge AI reduces the barriers for AI adoption by eliminating the need for continuous, costly cloud connections. In an exclusive interaction Rajesh Subramaniam, CEO & Founder, embedUR tells Rajneesh De, Group Editor, APAC Media & CXO Media that this enables AI deployment across small setups and ensures the true democratization of AI.
As India invests heavily in the IndiaAI Mission and builds compute infrastructure, foundation models, and national datasets, can the country truly achieve sovereign AI if intelligence remains concentrated in the cloud while deployment at the edge is overlooked?
Sovereignty is not solely about owning the model or data; it’s about controlling where and how intelligence is implemented. Even if India develops top-tier foundation models, if each inference must be sent to a centralized cloud, the actual decision-making would rely on connectivity, latency, and infrastructure that may be outside the country’s control.
Genuine sovereign AI means processing intelligence at the data source—whether on the factory floor, within hospitals, or at substations. Cloud and edge computing are complementary strategies; however, relying only on the cloud leaves sovereignty unfulfilled. True control over AI requires owning the last mile where it operates.
Why must India expand its focus from centralized data centres to sovereign edge infrastructure as AI moves into factories, hospitals, energy systems, and connected devices?
Data centres are crucial but weren’t designed for AI needs in physical environments. Once intelligence moves to the factory floor, hospital, or power grid, the limitations fundamentally change. Take latency: a machine making quality-control decisions on an assembly line can’t wait 200 milliseconds for cloud feedback — by then, the defective part is already past, or the machine hasn’t stopped. Detection must be real-time and local to be useful.
Reliability is another concern: a patient monitor relying on the cloud loses functionality if connectivity drops, and in critical care, those moments are vital. Life-critical systems can’t depend on network-based intelligence that might fail. Privacy and sovereignty issues also arise: routing sensitive data through external networks creates security risks and dependencies outside national control, which sovereign infrastructure aims to prevent. These challenges are inherently edge problems. Sovereign edge infrastructure—local, intelligent devices—provides India with resilience, privacy, and real-time response that centralized systems can’t match. As AI increasingly interacts with the physical world, this world operates at the edge.
How can Edge AI democratize AI adoption by bringing intelligence closer to where data is generated?
Edge AI reduces the barriers preventing most organizations from adopting AI. It eliminates the need for continuous, costly cloud connections, enabling AI deployment across India—from major metros to Tier 2 cities, rural clinics, and agricultural cooperatives. These entities can implement intelligent devices that operate locally without requiring a cloud infrastructure first. When AI resides on affordable, low-power devices near the data source, it shifts from being a privilege of well-funded companies to an accessible tool for anyone with a real problem. This exemplifies true democratization.
What role will embedded systems, edge computing, and intelligent devices play in determining India’s long-term AI success, and will these capabilities ultimately become the defining factor in achieving India’s sovereign AI ambitions?
Embedded systems represent the intersection of AI and the real economy. While foundation models grab headlines, their true value arises when intelligence is integrated into millions of devices powering India’s factories, utilities, and healthcare systems. India is well-positioned in this field, thanks to its strong engineering talent in embedded software and systems design. Moreover, the ability to deploy efficient, reliable AI on edge hardware might become the key competitive advantage — as it is the most challenging to perfect and the one that truly enables large-scale impact. Ultimately, a nation’s sovereignty will be judged not just by the models it develops, but by how effectively it can deploy and operate AI in practical, real-world environments.
Is India’s AI competitiveness at risk if innovation is not translated into scalable, real-world deployments beyond the cloud?
Yes. Innovation confined to labs or the cloud remains potential rather than competitiveness. Countries that succeed in the AI era will be those that transform research into large-scale deployed systems in the physical world. If India’s AI efforts stay limited to cloud models with slow deployment, it risks developing impressive capabilities that don’t lead to increased productivity, exports, or strategic gains. The difference between a prototype and a million reliably functioning devices outdoors is vast, and bridging that gap is ultimately what determines competitiveness.
How must national initiatives such as the IndiaAI Mission, Digital India, and Make in India converge around edge deployment capabilities?
These initiatives are natural partners but have so far largely operated in parallel. IndiaAI focuses on developing AI technology; Digital India provides the digital infrastructure, and Make in India creates hardware and manufacturing capabilities. Edge deployment is the point where all three converge: it needs indigenous AI, relies on digital infrastructure, and requires devices designed and manufactured in India. When these missions align around a common goal of deploying sovereign AI on Indian-made hardware, they strengthen each other. Edge AI acts as the connecting element that unifies these efforts into a single, comprehensive national capability.
What will it take for Indian enterprises across manufacturing, healthcare, and energy to convert AI innovation into scalable, real-world impact?
Three things. First, a shift in mindset — treating AI as an operational capability embedded into products and processes, not a pilot project bolted on afterward. Second, the right tools and partnerships to bridge the hard engineering gap between a model that works in testing and a fleet of devices that works reliably in production, at low power, in the field. Third, talent that understands both AI and the constraints of embedded, real-world systems. Indian enterprises have the ambition and increasingly the data; what they need is the deployment engineering to make it real. That’s the muscle we have to build collective effort, and it remains very achievable.
