Indigenous databases like ShaktiDB will reduce dependence on foreign interests: Shankar Raman, CEO, IITM Pravartak

An engineering graduate from Sri Venkateswara College of Engineering, Chennai, he holds an MS from IISc Bangalore and a PhD in computer systems networking and telecommunications from IIT Madras, Shankar Raman’s research interests include networking, internet energy efficiency, computer architecture, VLSI, and software testing.

Shankar Raman spoke to indianexpress.com on the research focus at IITM Pravartak and the impact it has had, their work with startups and their focus on ShaktiDB, an indigenous database. Edited excerpts:

Venkatesh Kannaiah: What is IITM Pravartak about and what does it seek to achieve?

Shankar Raman: The Indian government launched the National Mission on Interdisciplinary Cyber-Physical Systems (NMICPS) and under that mission 25 technology innovation hubs were established across premier institutions, including IIT Madras. We are one of those hubs, and we specialise in sensors, networking, actuators and control systems. These are core technologies that cut across many industries — from healthcare and agriculture to defence and transportation.

NMICPS was initiated because the government recognised a significant problem in the Indian research ecosystem. Public-sector R&D has traditionally been slow to respond to evolving tech needs, while private-sector investment in deep tech research remains very limited. The uncomfortable truth is that much of what we use or export is based on other nations’ technology stacks. Our domestic development of end-to-end systems has been weak. Around 80% of India’s R&D expenditure still comes from the government.

Our primary focus is on translational research; bridging the gap between academic research and societal or commercial application. Traditional research produces publications; translational research produces technologies that can be deployed on the ground.

Venkatesh Kannaiah: Can you give examples of the kind of work Pravartak is engaged in?

Shankar Raman: One example is in healthcare. We have supported the creation of a low-cost digital stethoscope that can detect cardiac dysfunction in children. This device is especially valuable in rural or remote areas where advanced diagnostic equipment like ultrasound machines may not be available. It allows early detection of congenital heart issues, so that the child can be referred from a village health centre to a district hospital for timely intervention. It’s affordable, portable, and designed for use in India’s primary healthcare network.

We also fund projects that serve broader social purposes indirectly. For example, TuTr Hyperloop aims to create ultra-high-speed, low-cost transportation for both passengers and goods. It has a potential to transform logistics and mobility at lower energy costs.

Venkatesh Kannaiah: What is the funding and sustainability model behind such initiatives?

Shankar Raman: Each Technology Innovation Hub received a Rs 170‑crore grant from the government for an initial five-year period. These hubs are not meant to operate indefinitely on grants. We were told to generate our own revenue streams, become self-sustaining, and if we fail, we would be shut down. In a way, each hub is like a startup. This kind of accountability is rare in government programmes. Out of the 25 original hubs, only about seven have reached a stable, sustainable stage.

The government’s intention was to create entities that blend social purpose with business discipline. You can’t rely on government grants forever; you have to create something that delivers value, earns revenue, and continues to innovate.

Venkatesh Kannaiah: What exactly is the mandate?

Shankar Raman: The government outlined four pillars for every hub: technology development, entrepreneurship support, skilling and training, and international collaboration.

In technology development, we identify promising research from across higher-education institutions beyond just IIT Madras, and help convert it into usable products or services. The goal is to take their academic R&D and translate it into something tangible.

In entrepreneurship, we support startups and innovators with seed funding — up to Rs 50 lakh — to convert prototypes into market-ready products. For some, we also provide additional follow-up funding when they transition from concept to company formation.

Skilling is an equally important pillar. Many graduates today are not immediately employable in advanced technology domains. We conduct deep-tech training for students, particularly from economically weaker sections and Tier‑2/Tier‑3 cities. They come to Pravartak for six months of hands-on training, receive a stipend, and get placed in companies that need such specialised skills.

Finally, we work on international collaboration. We bring foreign technologies to India and adapt them to our context or export Indian innovations abroad. For instance, we collaborate with Malaysia and Sri Lanka to conduct entrepreneurship and innovation development programmes.

Venkatesh Kannaiah: Could you talk about some research and product success stories?

Shankar Raman: In semiconductors and microprocessors one of our projects was with Mindgrove, which took the open-source “Shakti” microprocessor developed at IIT Madras and commercialised it into secure, production-grade chips. Their first chips came out of Taiwan’s TSMC foundry and are now being evaluated by various clients.

In aviation and avionics we have tied up with the Airports Authority of India and Airbus to run advanced air-safety management training. With India building new airports rapidly, pilot safety and operational management have become critical.

As for heritage and culture, we are using AI-driven digitisation tools. At Hampi, for instance, generative AI combined with old copper-plate inscriptions and photographs helped us reinterpret the famous Narasimha statue. Our model suggested it was originally Lakshmi Narasimha, and subsequent historical verification proved this correct. Earlier, this statue was thought to be of Yoga Narasimha.

We are also digitising fragile palm-leaf manuscripts using AI that is trained not only on visual cues but also on linguistic context, such as Saiva Agama grammar. This allows damaged words or letters to be intelligently reconstructed, something not possible with plain image recognition.

In the field of security and telecom there are startups like Folium Sensing that are deploying optical-fibre cables that act as sensors, detecting motion or vibration across large areas, useful for defence and perimeter security. Meanwhile, our 5G device security testing lab, built on IIT Madras’s earlier work on Indian 5G standards, is the only one in India that certifies 5G equipment for cyber-security compliance.

We also work in agri-tech. We feel that India’s shift to organic farming must be gradual. Our research shows that soil transition from inorganic to organic practices typically takes five years. We design farmer tools; battery-operated de‑weeders, portable weather stations to help them monitor soil and weather health to maintain income during the transition.

Venkatesh Kannaiah: How does AI contribute to heritage restoration practices?

Shankar Raman: The process is quite intricate. For palm-leaf manuscripts, we first digitise the content through high-resolution imaging. Then we use machine learning combined with domain-specific linguistic knowledge like Saiva Agama to fill missing portions contextually. This goes far beyond optical character recognition. Similarly, when reconstructing sculpture, we use generative models to simulate original forms based on multiple reference sources — inscriptions, photographs, even literary descriptions. Human experts then validate the AI output. It’s a marriage of art, archaeology, and advanced computing.

Venkatesh Kannaiah: You mentioned startups and incubation. How is Pravartak structured to nurture startups?

Shankar Raman: Pravartak doesn’t run a completely independent incubation cell, we work in close partnership with the IIT Madras Incubation Cell to ensure there’s no duplication or competition. Think of it as a continuum. Pravartak serves as the pre-incubation stage, where innovators come in with a proof of concept or a research idea and receive their first structured support. Once they develop a prototype and gain some early traction, they transition to the IITM Incubation Cell.

In our pre-incubation phase, we offer several key supports. First, financial support — up to Rs 50 lakh per startup. This funding helps them build a prototype or move from concept to product. Second, physical space — we provide dedicated work areas. Third, we connect them with mentors who can help them refine their product-market fit. Fourth, we link them with industry partners, investors, and pilot customers.

Additionally, to ensure accountability and long-term engagement, we take a small equity stake, typically 5%, in every startup we fund. Importantly, this is not taken as a condition of the grant; we purchase it separately at par value.

Venkatesh Kannaiah: How many startups do you currently support, and could you highlight a few examples?

Shankar Raman: As of now, we have supported 52 startups through this model, and 13 have already graduated successfully. Each one represents a different aspect of deep technology and social impact.

For instance, Music Temple has created a formal written notation for Indian classical and folk music, something that never existed before. Mindgrove is developing India’s first commercial RISC‑V microprocessor products under the Shakti architecture. TuTr Hyperloop is working on sustainable, low-cost, high-speed transportation solutions. Folium Sensing has built optical fibre-based sensors capable of detecting motion, stress, and vibrations — vital for defence and infrastructure monitoring.

In healthcare, several startups are working on low-cost diagnostic devices that can operate even in resource-limited environments. Many of these are already in pilot use at district hospitals and primary healthcare centres.

Venkatesh Kannaiah: Tell us about ShaktiDB and why is it a critical project?

Shankar Raman: Databases are the foundation of all digital systems. Every application: AI, analytics, finance, or defence relies on one. Currently, India depends heavily on foreign databases like Oracle, Microsoft SQL Server, or MySQL. That dependency introduces both vulnerability and cost. Suppose geopolitical tensions arise or licenses are withdrawn — entire systems could be affected. We saw a similar pattern globally, where even US agencies have moved away from commercial databases to their own secure variants. China, for instance, developed indigenous systems used by Alibaba and others.

That’s why we proposed Shakti DB. It is based on PostgreSQL, an open-source database, but we have enhanced it with three India-specific elements. We have built-in regulatory compliance features for RBI, SEBI and other agencies, we have added strong security layers designed for sensitive data environments, and we are mentoring local support and maintenance through Indian service providers. The Ministry of Electronics and IT (MeitY) saw its national relevance and funded the project.

Our approach is not just to build software and leave it there. We are building an ecosystem. We have incubated six support companies whose sole purpose is to maintain, customise and deploy ShaktiDB for clients. These firms will provide on-the-ground support, something foreign database vendors rarely do here. We already have pilot users and are collaborating with companies like Zoho to explore integration. Over the next five years, we expect ShaktiDB to mature into a secure, homegrown alternative that can power India’s digital infrastructure.

Venkatesh Kannaiah: Building something like that sounds expensive.

Shankar Raman: Yes. Building and testing databases requires substantial compute resources. But instead of reinventing the wheel, we leverage national institutions that already have such infrastructure. We partner with RailTel for networking support, CDAC for compute power, and STPI for innovation ecosystem support. Pravartak’s job is to be the catalyst — to spark collaboration, connect the right players, and manage the mission efficiently.

With the right partnerships and frugal engineering principles, Indian projects can achieve global-class results at a fraction of the cost.

Venkatesh Kannaiah: There is a grouse that private-sector investment in R&D is low. How does it reflect on the ground?

Shankar Raman: In India, unfortunately, the private sector has often prioritised short-term gains and outsourcing models over genuine R&D. As a result, we have become world-class service providers but not product innovators. Contrast that with sectors like automotive or aerospace, where private R&D investment is substantial and you will see India has achieved much more autonomy there. In IT and deep tech, however, the lack of sustained private research has left us dependent on global suppliers.

What we’re doing at Pravartak is to change that pattern — at least for a few high-impact domains. We de-risk early-stage research and development so that private industry can later step in and scale it commercially.

Venkatesh Kannaiah: What is the single biggest problem you’re trying to solve?

Shankar Raman: Indigenisation. In simple terms, ensuring that India can design, manufacture, and maintain critical technologies without relying on other countries. Whether it’s a secure database like ShaktiDB, semiconductor chips, or healthcare devices, the idea is the same self-reliance with world-class quality. Every imported product carries a dependency risk and an invisible cost. We want to reduce both.

At the same time, cost-effectiveness matters. Technologies built in India must be affordable for Indian users whether for farmers, hospitals or defence establishments. Security and cost are the twin pillars we balance in every innovation we pursue.