‘Our data, our models’: The push for AI sovereignty

The transformative impact of cloud and AI technologies on India's economy is becoming increasingly evident, particularly in the context of inclusive growth and the need for local data processing to address the country’s unique challenges. In a conversation with Voice&Data, Swastik Chakraborty, VP–Technology at Netweb Technologies, highlighted the expected doubling of India’s data center capacity by 2027. He also underscored the critical importance of data sovereignty, governance, and lineage in effectively managing the rapidly growing volume of data.

The conversation also explored the applications of AI in various sectors, particularly banking, finance, and healthcare. Swastik detailed how AI enhances fraud detection and loan approvals in financial services, while also improving operational efficiency in telcos and banks, and much more. Have a look at the excerpts from the interaction:

How will cloud and AI infrastructure define India's role in the global economy?

If you look at the perspective the Government of India is adopting as part of this overall digital transformation journey, there are several key focus areas. The foremost among them is inclusive growth. Historically, not all communities in India have been able to participate equally in driving the national economy. Access to technology has often been limited, due to cost, infrastructure, and other barriers.

Cloud infrastructure helps bridge that gap by enabling broader access to IT resources. AI, traditionally a cost-prohibitive technology, is now being made more accessible. Under this inclusive model, someone with a bright idea can now experiment quickly, and if the prototype succeeds, it can move rapidly into implementation.

So, the proliferation of cloud has made tech consumption more affordable and available across demographics. AI is now helping to accelerate the journey from ideation to experimentation and implementation. That puzzle, accessibility, affordability, and scalability, is being solved. That’s also why the Government is using these technologies to boost not only employment but also workforce productivity.

That, in my view, is the role AI and cloud are playing in shaping India’s economic future.

India’s data centre capacity is expected to double by 2027. What’s driving this rapid expansion?

Absolutely. As India becomes a more vibrant economic power, with a growing GDP and dynamic startup ecosystem, this expansion is being fueled by homegrown innovation.

India’s problems are unique. Our data types, infrastructure challenges, and user behaviours differ vastly from the global norm. So, naturally, our solutions must also be uniquely Indian. That’s where data infrastructure and AI democratisation come into play.

To create AI systems that truly reflect Indian needs, we need massive amounts of localised data and computing power. And that’s why the development of foundation models, built from scratch using Indian data, is so crucial.

This, in turn, calls for significant computing power, which brings us back to the growth in data centres. As we democratise AI and build foundational models tailored for Indian use cases, we need a decentralised network of high-performance data centres, even at the edge.

These data centres must also scale in terms of energy infrastructure, cooling, and network connectivity to support this demand. And that's what we’re seeing: a data infrastructure revolution that supports IT and AI expansion across India.

Since you mentioned AI, do you believe this data centre growth will impact cloud adoption and data localisation?

Absolutely. As I often say, India’s data is India’s responsibility. Our data reflects our culture, behaviours, and socio-economic realities, very different from those elsewhere.

That’s why data sovereignty is so vital. If I ask a question to an AI system, it shouldn't need to leave the country for processing and then return with an inference. We need both data localisation and model localisation.

We must ensure our data and models remain within national boundaries. This is critical not just for sovereignty but also for building responsible, sustainable AI that truly serves Indian citizens. It’s not just about where the data lives, it’s about where and how it’s processed.

With more data centres being built, are storage costs coming down and access speeds improving for businesses and users?

Traditionally, humans and IT systems generated data. Today, we’re also seeing an explosion of synthetic data, especially from AI applications. In the next two to three years, data volumes are expected to double, and that creates both a challenge and an opportunity for storage.

We’ll need high-capacity, high-performance storage systems to house these enormous data corpora, especially for training AI models. But it's not just about capacity or cost-per-gigabyte. The real challenge lies in data governance and lineage.

Where did the data come from? Is it authentic? Who has authorised its use? We need strong systems in place for data traceability and compliance.

Think of storage as the container, whether you pour water or oil, the content must be trustworthy and governed. So while we can reduce storage costs with better technologies, governance and lineage are the true priorities now.

You earlier mentioned, “Our data, our model.” Do you think national programmes like Digital India and India AI Mission are shaping cloud and AI infrastructure adoption?

The India AI Mission is a cornerstone of this transformation, and I commend the Government for turning this vision into reality. There are four major pillars that I believe define this mission.

The first is Infrastructure Enablement. The Government is establishing a two-sided ecosystem comprising both providers and consumers, which lies at the heart of AI democratisation. This ensures that AI technologies are not only developed but also made accessible and usable across various sectors of society.

The second pillar is the Data Ecosystem. Through initiatives like the IndiaAI Datasets Platform and the AI course, the Government is actively working to build a robust and scalable data backbone. These efforts are essential for training high-quality AI models and encouraging innovation grounded in Indian data.

Third is the focus on Foundation Models, particularly those trained on Indian data for Indian use cases. One compelling example is the preservation of cultural heritage, such as Bharatnatyam, through AI models. This not only safeguards our traditions but also makes them more accessible to future generations.

The fourth and equally important pillar is Talent Development. By integrating AI into school, college, and university curricula, we are laying the foundation for a future-ready workforce. This initiative aims to cultivate the next generation of AI innovators and practitioners from an early stage.

Together, these pillars form a continuous innovation loop, from ideation to experimentation to implementation, all fuelled by the convergence of cloud technologies and AI.

AI is transforming banking and finance in India. How is it helping with fraud detection and other critical use cases?

AI has a huge role in fraud detection, loan approvals, and insurance verification.

Let’s take fraud detection. AI is helping reduce false positives (flagging something as fraud when it isn’t) and false negatives (missing actual fraud). Both are costly—one wastes time; the other is dangerous.

By using graph-based AI models, banks can better understand relationships between transactions and flag anomalies more accurately. Similarly, in insurance, AI can prevent multiple claims from the same individual across providers.

At its core, AI is boosting employee productivity by automating repetitive, error-prone tasks and letting human experts focus on high-value work.

What about healthcare? Are Indian hospitals using cloud-based AI for more accurate diagnosis and better delivery?

For the past six months, I’ve been working on exactly this. Under the Ayushman Bharat scheme, we have Public Health Centres (PHCs) across India. But many PHCs lack available doctors, delaying treatment. So we built an AI-powered Doctor Assistant, not to replace doctors, but to triage patients faster.

Patients can talk to this AI assistant in their local language, Bengali, Marathi, Hindi, Kannada, and more. The system then relays a summarised version in English to the doctor, who can remotely prescribe tests or treatment.

We also developed a Radiologist Assistant that pre-screens scans and suggests possible diagnoses, which the radiologist can confirm or modify.

We're working to integrate this with ABHA (Ayushman Bharat Health Account) so it connects to government EMR/EHR systems. These agents significantly improve doctor and patient experiences, especially in remote areas.

With startups booming in India, how are they using cloud and AI tools to grow and compete globally?

Startups are adopting AI at scale for one reason: accessibility. Today, powerful AI tools are available at an affordable cost.

We used to talk about the "idea economy",now we have the tools to act on those ideas. AI is no longer an afterthought; it’s being embedded into the fabric of startup strategies.

However, not all challenges are solved. AI models need guardrails, bias checks, and security protocols. For example, if model parameters are tampered with, the system could produce flawed outcomes.

Startups are moving fast, but we also need to focus on maintainability, accuracy, and security to ensure long-term impact.

The democratisation of AI is one of the biggest drivers of employment creation today. There’s often a misconception that AI will lead to job losses, but that’s absolutely not the case. In fact, bringing AI into the hands of more people is a key strategy even for the Government of India to generate employment opportunities.

Think of it this way: earlier, IT was largely the domain of business managers. But today, even teams like sales and procurement are asking whether they can use AI models to boost their productivity. For example, in procurement, people are exploring AI tools that could assist in evaluating RFPs more efficiently.

If you look at the BFSI sector, one of the key challenges is document verification. When someone applies for a loan, they often need to submit 30–40 documents. Who verifies them, and how long does that take? AI can help boost the productivity of the processing staff, someone who might currently process one loan per day could potentially handle ten or even twenty.

This opens up opportunities for startups to develop AI solutions that enhance productivity. That, in itself, is job creation, not only for those building the tech, but for those benefiting from improved business processes.

Ultimately, I believe AI enables people to build solutions, develop new business models, and drive measurable productivity gains. That’s the true business value we should focus on.

In terms of indigenous tech development, how crucial is self-reliance in semiconductors for India’s broader digital economy and national security?

Before we run, we need to learn to crawl and walk, and I think that’s the stage we’re at right now.

India is making deliberate moves to position itself as a semiconductor powerhouse. And we need to start somewhere. Companies like NetWeb, for instance, have begun manufacturing indigenous motherboards, reducing dependency in at least one critical area.

But semiconductors encompass much more than just processors. We’re talking about GPUs, memory modules, storage like NAND chips, and so on. The Government of India is well aware of these needs, and there have been recent developments on that front.

Self-reliance is, of course, the long-term goal. But in the short term, it's about being able to meet the soaring demand from the data centre industry. Today, businesses expect instant delivery and deployment. Waiting a month or two just isn’t feasible.

Beyond the chips themselves, we also need to build capabilities in systems integration. You might have a processor or GPU, but if a specific register or memory module is missing, your server won’t function. Every component plays a role in the value chain.

So, it’s not only about manufacturing processors or GPUs, it’s about developing a holistic ecosystem for system-level manufacturing. We are currently in the crawling phase, but we’re moving steadily towards walking, and eventually running. The important thing is, we’ve started the journey.

How do you think India can enhance its global competitiveness in building advanced cloud and AI infrastructure, particularly when supported by domestic innovation, like indigenous semiconductor chips or homegrown data centres?

There’s already a lot of innovation coming out of India. If you look at patent filings over the last three to five years, there's a noticeable uptick. That’s a strong indicator of intellectual and technological progress.

In our own work on server development, for instance, we’ve been introducing new technologies, sometimes aligned with global standards, and sometimes even leading the way.

Government support, along with collaboration between industry and academia, is essential to accelerate this momentum. We are at a point of significant technological transition, and newer innovations are reshaping what a data centre looks like.

Personally, I come from a research background, I used to work with the Department of Space, so I’ve seen first-hand how deeply India is invested in research and new technologies. That spirit continues today.

At NetWeb, we're contributing through innovations in storage, HPC (High-Performance Computing), and AI infrastructure. These are the areas where we're trying to make a tangible impact.

In your view, is India on track to becoming a global hub for advanced technologies such as data centres, AI, and cloud? And what more needs to be done?

I believe we are firmly on the path to becoming a global provider, not just of data centres, but of the solutions that run on them.

India's domestic demand is already huge, and rightly, that should be our first priority. But once we meet internal demand, the next step is “India for the world”.

We already have many of the key elements in place, government support, industry backing, and increasingly strong collaboration between research institutions and the private sector.

What’s needed now is to bridge the gap between research and market-ready solutions. We must focus more on converting research outcomes into real-world products and business value.

Looking ahead, we should continue nurturing innovation, whether to solve India’s unique challenges or to address global problems. In the short term, the emphasis should be on strengthening research, industry partnerships to bring commercially viable solutions to market.

And with the government’s continued support, I believe we’re on the right track to achieving that goal.