India can build its own AI cloud, reduce reliance on foreign compute power, says RackBank CEO

The data center and Artificial Intelligence (AI) infrastructure industry is growing fast, driven by demand for powerful, efficient, and sustainable systems across sectors like healthcare and finance. RackBank, a key player in this space, focuses on building data center infrastructure tailored for AI workloads in India that deploy faster.  

In a conversation with TechCircle, Narendra Sen, Founder and CEO of RackBank shared insights into RackBank’s AI-focused expansion, the development of their AI Super Cloud platform called NEEV, and their vision for India to become a global leader in AI infrastructure. He discussed how the company’s integrated approach, from hardware to cloud orchestration, aims to support Indian innovators and startups while addressing sustainability and geopolitical challenges. Edited Excerpts: 

What inspired your recent expansion in AI-focused data center infrastructure, and how does it fit into your long-term vision for the company?

We’re focused on building AI infrastructure and the capabilities required to support it in India. India has strong technical talent, and there's no reason we can't also build large language models (LLMs) here or become a global player in AI infrastructure. 

The foundation of AI is digital infrastructure. Without it, it's impossible to build systems or agents that operate continuously and process large volumes of data in real time. 

This kind of infrastructure needs specialised compute power, like GPUs, and has different requirements than traditional data centers. We've redesigned our data centers specifically for AI workloads. Unlike traditional setups that take 24 months to build, our AI-optimized, horizontal data centers can be built in 6–8 months. Everything is pre-manufactured and assembled on-site, which speeds up deployment. 

To manage high power demands and reduce environmental impact, we use liquid cooling. This reduces water usage by 95% and cooling energy needs by 40–50%. We're also using clean energy, which is more affordable than grid power in India. These choices help enterprises adopt AI faster while keeping sustainability in mind. 

Our goal is to make AI infrastructure both available and affordable. In a cost-sensitive market like India, easy access to infrastructure removes barriers for developers and startups to innovate. 

In the past, Indian researchers often had to rely on US companies for compute power, sharing their work and patents. With our infrastructure in place, that’s no longer necessary. 

We're now building an AI "supercloud" on top of this infrastructure, vertically integrated with our own data centers. We call this product NEEV, which means "foundation" in Hindi. 

Since many global cloud platforms are built by Indian engineers, there’s no reason we can’t build and operate one ourselves. It’s not about the lack of technical ability, just the lack of initiative. So we’ve taken that step to build India’s own AI cloud company. 

How is your company positioning itself amid the growing demand for compute power driven by AI and competition from startups? 

The foundation of AI is digital infrastructure, which is our core area of expertise. Our focus is on making this infrastructure affordable and accessible, and that shapes how we approach everything, from selecting data center locations to optimising the orchestration engine, cloud layer, and hardware procurement. It's a fully integrated approach. 

When we began working on AI infrastructure, we realised the cooling requirements were substantial. To address this, we developed a liquid immersion cooling solution called Varuna. Our servers are submerged in a liquid coolant, which efficiently removes all the heat. 

We’ve built our own products with a focus on efficiency and innovation. Technology and innovation are the main ways to differentiate, and our entire focus is on AI infrastructure. We’re not building general-purpose platforms or applications and not entering unrelated areas. 

Currently, about 90% of our digital needs depend on foreign cloud providers. If that relationship were disrupted, it could impact our entire digital economy. We know this shift is difficult, but we’re confident in our ability to compete globally because of the country’s strong pool of engineering talent. 

What does your roadmap for infrastructure, scalability, and adaptability in India look like? 

We’re building AI-focused cloud and supercomputing infrastructure, what we call the AI Super Cloud. This system is designed to provide access to high-performance compute for researchers and developers working on AI innovation. 

AI is already accelerating progress in key areas. For example, drug discovery timelines are being reduced from eight years to two. Many problems that have remained unsolved for decades, especially in healthcare, can potentially be addressed through AI, but this requires the right infrastructure and platforms. 

Our goal is to support Indian companies with this infrastructure, not just within India but globally, similar to how AWS helped scale US companies. We want to do the same for Indian SaaS and AI companies, enabling them to train and deploy models at scale. 

We’ll continue building products that let companies focus on innovation rather than infrastructure. Because in AI, speed and scale are critical. 

There’s also a geopolitical angle. The US is shaping global AI policy and restricting chip access. India has the opportunity to counterbalance this by building open, scalable AI platforms that can serve other nations, just as UPI changed the payments landscape and challenged legacy systems like credit card networks. 

AI can be democratised, and India is well-positioned to lead that effort. Our market includes critical use cases in healthcare, drug discovery, urban safety, and finance. All of these require powerful compute infrastructure, and that’s what we’re building. 

Beyond liquid immersion cooling, what other technologies is your company exploring to reduce environmental impact and improve energy efficiency in data centers? 

We were early in focusing on sustainability. Securing clean power is something many can do, but we’ve built a purpose-designed liquid-cooled data center. 

You may have heard about Microsoft submerging servers in the ocean. Our approach improves on that, because when servers are underwater, they can’t be easily maintained. With liquid cooling, systems stay accessible and manageable. Also, India has a hot and humid climate, and water is needed for people, not machines, so we designed accordingly. 

We started by focusing on clean energy and minimizing water usage. We follow key efficiency metrics: PUE (power usage effectiveness), WUE (water usage effectiveness), and CUE (carbon usage effectiveness). Our goal is full end-to-end optimization toward net-zero. 

We also considered building materials. Instead of cement, we use steel structures to reduce carbon emissions. Our locations are selected for environmental support, areas where we get 6–8 months of free cooling without running heavy cooling systems. 

We're also exploring regions like Northeast and North India, places like Jammu & Kashmir and Assam, where we can get natural cooling for 10–12 months a year. This helps reduce reliance on mechanical cooling systems. 

With so many technologies and so much data in play, which ones, like AI, cloud, or green data infrastructure, excite you most about the future? And where do you see all this heading? 

AI is transforming industries by enabling machines to think and act like humans. One example is how AI systems can now play complex games like Jeopardy, showing human-like reasoning. 

We're witnessing rapid innovation. In areas like robotics and automation, factories are beginning to use AI agents at scale. In India, these agents are being deployed widely, even in remote areas. 

Programs like Safe City are already using AI, including AI-generated calls for insurance. These are early days for AI adoption, but the impact is expected to grow significantly. Over the next five years, AI could boost productivity by 100 times, shift focus away from repetitive tasks, and improve work quality. This will also contribute to India's GDP. With current targets around $4–5 trillion, AI could help double that. 

To support this, the Indian government is pushing initiatives under the India AI mission, including efforts to make GPUs more affordable. The government is proactive and aligned with global AI trends. The ecosystem is also responding, accelerating adoption across sectors. 

India is well-positioned to be a global leader in AI.