From ancient temples to AI Chips, inside IITM Pravartak’s deep-tech playbook. Backed by IIT Madras and DST, IITM Pravartak is driving research into real-world impact—from AI-led heritage restoration to secure embedded systems. In this interview, Shankar Raman tells EFY’s Akanksha Sondhi Gaur how our bold incubation model is fuelling India’s next tech leap.
Q. What sets Pravartak’s incubation model apart, and how do you measure its success?
A. We go beyond funding and space. We actively nurture startups by embedding with them, guiding them daily for up to two years. We focus on both cutting-edge deep-tech and scalable ideas from modest R\&D. Every incubatee must validate market fit through the GDC cohort programme. Projects are co-developed with IIT faculty and industry experts, ensuring real-world relevance. Our model is hands-on, with strong customer and VC connections, treating startups as partners, not tenants. It is not about how many companies we incubate; it is about how many are still standing five years later. That is the true measure of impact. At Pravartak, our focus is on building that survivability through funding strategies, customer connects, and business model refinement.
Q. What is your 3–5 year roadmap and long-term vision for India in electronics and deep-tech?
A. We are focusing on two major areas: aviation and electronics, with a special emphasis on safety and cybersecurity. In aviation, the goal is to promote safe aviation engineering. In electronics, we are zeroing in on secure components and embedded cybersecurity. Given the rise in cyber threats and the strategic importance of electronics, this is a natural direction for us. We need to create opportunities at home for our talent, not just be a brain-exporting nation. Our focus must be on solving India-specific problems with India-developed technologies. Over time, we can improve quality and scale globally. With government support, industry collaboration, and focused translational R\&D, India can become a true deep-tech powerhouse.
Q. How do local needs drive innovation at IITM Pravartak?
A. India’s context is fundamentally different. For instance, 5G was originally designed for high-speed transport in Western countries, like cars zipping at 300 km/h on autobahns. Whereas in India, most people travel by autos, two-wheelers, or bicycles at 50–100 km/h. So why import that model blindly? Our approach is to develop India-specific, cost-effective solutions tailored to the country’s needs. Local problems demand local innovation—and that is not “jugaad”; it is intelligent adaptation. For example, we are working on low-cost digital telescopes for cardiac diagnostics. Instead of the standard ₹85,000+ systems, we are targeting ₹50,000 devices with acceptable error margins that still ensure no critical condition goes undetected.
Q. How do you support startups beyond incubation?
A. If startups lack business or marketing capability, we assign in-house experts to represent their product—but for a fee. It is a nominal charge, but it is important. We want them to understand that every service has value, and in business, financial discipline is vital. We are not a subsidy body—we are building responsible entrepreneurs.
Q. How do you choose innovations to incubate and turn them into products, especially in AI, IoT, and cybersecurity?
A. We use a two-stage filter: a technical committee vets the idea’s feasibility, and a business panel assesses market potential. Only ideas with both depth and viability move ahead. For example, in AI, we helped develop multilingual translation tools by adding domain-specific context—solving issues generic models miss, like misinterpreting ‘I tabled a bill.’ This led to our Centre for Human-Centric AI, focused on making compact, deployable models for real-world use in areas such as edge AI and robotics.
Q. What founder traits do you seek, and how do you balance tech and market fit?
A. Passion, adaptability, and customer empathy. We value founders who persist through failure, are open to changing their core idea based on feedback, and truly listen to customers—traits more critical than the tech itself. Programmes such as GDC training reinforce this mindset.
We address both ends. First, we use an automated software tool to assess patentability or detect existing IP conflicts. Then, a team comprising industry and academic experts evaluates the application. Sometimes, we have face-to-face discussions to understand their conviction better. We do not rely heavily on theoretical market projections. Many startup pitch decks forecast revenue for 2025 or beyond, but to us, that is just paper. What matters is: Can your product genuinely solve a customer’s problem? If yes, the business will grow organically.
Q. How do you support startups beyond incubation, especially to cross the valley of death?
A. That phase, typically in years 3 to 5, is where many startups fail due to a lack of cash flow. We have learned from IIT Madras Research Park that the key is to ensure cash inflow by Year 2. We do this by facilitating early service contracts or pilot deployments. Once cash starts flowing, their chances of survival increase significantly.
Q. How do you nurture student-led startups and tap early IIT ideas?
A. Yes, we actively nurture student-led startups, especially those emerging from IIT Madras’s Centre for Innovation (CFI). CFI is a hub where students build proof-of-concept projects. We scout these early ideas, and if they show promise and market potential, we bridge them to customers or bring them under Pravartak or the IITM Research Park for incubation.While we don’t incubate all sectors—like healthcare or rural tech, we focus heavily on deep-tech domains such as sensors, actuators, networking, and control systems. Startups like Folium Sensing (optical sensors), Rio Systems (wireless sensing), and others working on electrical wave-based sensing are examples we have supported.
We offer strategic handholding at every stage, from concept to commercialisation, ensuring that innovation doesn’t just stay in the lab. For sectors that require costly tools (like EDA software), we negotiate bulk licenses with vendors at discounted rates. Startups gain early access they would otherwise not be able to afford, and vendors benefit when those startups scale into paying customers. It is a pipeline built on collaboration, context, and real-world application.
Q. Could you share examples where research incubated turned into commercial success?
A. A great example of our deep-tech incubation success is TuTr Hyperloop, born out of the student-led Avishkar Hyperloop project, which recently secured its first customer, Karaikal Port. Another is Mr Jan’s startup from IIITDM Kancheepuram, which has developed an electrical wave-based device to assess concrete strength and is already in the market. MindGrow, focused on affordable electronics manufacturing, is actively engaging with potential customers. While deep-tech startups usually take 5–6 years to mature, and Pravartak itself is just over four years old, we are starting to see solid traction. The model works when startups, researchers, and industry align with intent and persistence.
Q. How do you choose and prioritise R&D focus areas in electronics and cyber-physical systems?
A. Our domain was defined through the National Mission on Interdisciplinary Cyber-Physical Systems (NMICPS) by the Department of Science & Technology (DST). We were assigned to focus on sensors, networking, actuators, and control systems, all critical building blocks of any cyber-physical system, be it drones, autonomous vehicles, or industrial IoT. This focus gave us both clarity and flexibility. We didn’t restrict ourselves to sectors like healthcare or automotive; instead, we looked at cross-cutting technologies that power multiple applications. That helped us build depth and scalability in foundational domains.
Q. How do you assess ROI and market readiness for projects like Hyperloop or secure chips?
A. We apply deep business thinking. Take Hyperloop as an example. We asked why Elon Musk abandoned it. The answer: passenger transport is a loss-making model, burdened with high safety and regulatory costs. We reimagined it for goods transport, which has fewer regulations and better margins. That pivot makes the technology a viable and profitable solution-led thinking over hype-driven models.
Q. How do you ensure research meets real-world needs in areas like defense or healthcare? Any examples?
A. We start by deeply understanding stakeholder pain points, be it from government or industry and then align them with IIT Madras’ research strengths. For example, in a heritage restoration project at Hampi, instead of relying solely on tech like shape reconstruction and material analysis, we worked with historians and used palm-leaf manuscripts, copper plates, and old photos. This blend of academic rigor and contextual insight ensures practical, culturally sensitive solutions.
Q. How key are indigenous platforms and IP in your semiconductor R&D?
A. Indigenous development is critical. Take the Shakti processor, an IIT Madras initiative, we helped incubate Mindgro, a startup that built on Shakti. This occurred even before government schemes like DLI or PLI were implemented. The semiconductor space remains tough. Without a domestic fabrication facility, chip design efforts will continue to face hurdles. When we started this journey in 2019–2020, there was no fab in India. But with recent government support, momentum is picking up.
Q. How do you support hardware startups? Any early challenges in prototyping or validation?
A. We offer EDA tools, connect startups with early customers, and provide board design and prototyping support through IIT Madras and global partners, such as TSMC, via MPW slots. Given the high costs, we use vendor tie-ups, MOUs with institutions, and shared infrastructure to ease access. We offer startups access to advanced labs, prototyping tools, and early validation through industry tie-ups—especially vital for ventures involving sensors or embedded systems. Earlier, hardware development was limited by ecosystem gaps, like a lack of access to high-speed ASICs and reliance on outdated components. Even with strong design skills, projects like the Shakti processor faced performance limitations due to subpar IP cores.
Q. Despite constraints, how did you gain momentum in semiconductor and component integration?
We were bold enough to take the plunge despite these limitations. We formed companies around our innovations. Over time, government support played a crucial role, it encouraged the growth of companies that could handle semiconductor design and even component-level integration. That is how we have built a thriving ecosystem around deep-tech electronics.
Q. How do you collaborate with startups and industry to ensure your solutions are deployment-ready?
A. We follow a dual-track approach. First, we leverage the strength of startups incubated at IIT Madras, the Research Park, and Pravartak. When a customer approaches us with a specific requirement, like defence needing portable medical machines, we evaluate our startup ecosystem and connect the right players. In one case, we linked two startups to co-develop a solution. The second approach is when the requirement demands new research. In such cases, we initiate collaborative projects through IIT Madras, particularly via IC&SR. We know which professors specialise in what areas, and we bring them into the loop. Once a solution is developed, our team validates it and then delivers it to the customer.
Q. How do you collaborate with startups, large enterprises, and global tech players?
A. We co-develop solutions with startups and industry leaders, not just offer services. Our model is built on joint IP and deep tech collaboration. Public partnerships include Accenture (CoE), Samsung Innovation Campus, Sony, and government agencies like DST, MoD, Supreme Court, and Sansad TV. We are also in advanced talks with major Indian industrial groups. All MoUs are shared on our website and social media. Our principle: “For every operation you do, we have technology.”
Q. What is your strategy to scale India-made innovations globally, particularly for markets with similar constraints?
A. India has two key advantages: cost and quality. From motorcycles to medical tech, we consistently deliver high-quality goods at competitive prices. We also have a huge domestic market, which helps us refine products. If something succeeds in India, it has a high chance of succeeding globally. Experience, scale, and affordability are the pillars of our go-global strategy.
Q. How does India’s scale offer an advantage in R&D, and what is needed to turn it into a global deep-tech hub?
A. India’s vast scale gives it a clear edge, as seen in a virtual autopsy project where a European firm had 3000 cadavers, but AIIMS could generate that in a week. This accelerates testing and innovation. However, to become a global R&D hub in electronics, India must go beyond talent. We lack core infrastructure, especially in semiconductor equipment. To lead in next-gen electronics, we need full-stack capability, from design to fabrication tools, with a focused, mission-mode push.
Q. How do you balance government grants with private funding, and what is the road ahead?
A. There is often a disconnect; industry funds R&D for real outcomes, but academia often stops at papers. That’s why under the NM-CPS mission, we focus on translational research, taking tech from lab to market. Our model blends DST grants with matching private/public funding, ensuring relevance and accountability. With ₹3 billion already raised and financial independence achieved, we are now eyeing international expansion via IITM Global, taking startups and tech abroad, not just to export, but to collaborate. Global scale brings new challenges, IP battles, cross-border pitching, patent quality. So we’re ramping up patent activity and training startups to navigate global markets, not just chasing unicorns, but building surprise innovations with deep impact.
Q. How have DST and private ties shaped your growth and partnerships?
A. Immensely. Both sectors are becoming more receptive to adopting new technologies. While private industry moves quickly, we see more resistance in government due to job security concerns. That’s why we also engage with policymakers, emphasising that technology won’t eliminate jobs but redefine roles. Our strategies differ accordingly for the public and private sectors, and we are seeing good progress on both fronts. It decentralises innovation and democratizes opportunity. By involving diverse institutions—rural colleges, private universities, and premier IITs, we are solving real problems with a broader talent base. It is not just about elite research; it is about inclusive problem-solving rooted in India’s unique needs.
Q. What policy shifts can boost deep-tech, and how can firms co-develop or license tech globally?
A. Inter-ministerial coordination is key. We often see contradictions between policies from the Ministry of Finance, Commerce, and Science & Technology. For example, deep tech R&D is currently governed by the same financial norms as road construction—this is impractical. Streamlined, responsive decision-making is crucial. If India doesn’t act fast, we risk seeing core technologies developed here but scaled abroad. We’ve prevented a few such exits, but many have slipped through.They can approach us directly or via IIT Madras. If it aligns with our focus, we are open to co-developing or licensing technologies. We are already working with multiple Fortune 500 firms and actively signing MoUs with four to five new players.
Q. Do startups rely too much on government support? What’s your take?
That’s a fair observation. Many times, I have seen private players ask, “What will the government give me?” While some support is reasonable, you can’t expect everything to be free and still aim for profit. At Pravartak, we make it a point to teach entrepreneurs to value what they receive and build sustainable models rather than rely on subsidies.
Q. How are emerging techs like quantum, neuromorphic, and secure chips shaping India’s future?
A. We’re currently not working in quantum, as it is managed under a different national mission. For other areas, we focus on problem-driven innovation. We begin with a customer’s problem statement, then identify the appropriate technology using IITs’ research capabilities.Our strength lies in the Pravartak team’s ability to translate complex tech into actionable solutions, thanks to our mix of academic experts and industry veterans. As Prof. Kamakoti often says, “You don’t need a gun to shoot a mosquito.” The tech applied must be right-sized, efficient, and relevant.
Q. What drove the use of GenAI in heritage work, and what does it reflect about your approach?
A. When we began examining restoration challenges, we asked: why not utilise AI or generative AI to reconstruct broken artefacts? Instead of just an engineering view, we took a customer-centric perspective. Customers often care more about the description of the object than the precise engineering details. Historically, poets described kings as ‘lion-like’—not literally, but to convey qualities. Similarly, many idols and statues have rich textual descriptions, and sometimes old photographs.
Using both, along with AI-driven triangulation, we can scientifically reconstruct an artifact’s original form—not just an artistic guess, but a proof-based recreation.It illustrates our core approach: don’t just solve for the problem—solve for the context. Research gives you precision, but application needs perspective. That’s what makes our model unique: we balance mathematical rigor with real-world insight.
Q. What’s the tech architecture behind the real-time translation in Parliament?
A. We built the system for Sansad TV with a human-in-the-loop model to handle name mispronunciations and prevent context errors—critical in formal settings. AI learns from human corrections over time. Major hurdles included dialect variation and ensuring translations stay respectful.
Q. How does your shared IP and tool licensing model work? Any regulatory hurdles?
A. We centrally license tools and share them with startups—cutting costs while ensuring ethical use and vendor trust. As a Section 8 non-profit, Pravartak focuses on social impact, not profit. It supports open-source and commercial IP through legal safeguards. But yes, hurdles like foreign payments and procurement policies need government help.
Q. How do you align skill development programs with startup and industry needs—especially in electronics and embedded systems?
A. The core issue is the declining quality of engineering education. Out of 10,000 hackathon ideas, maybe 0.1% are truly innovative. Today’s engineers often lack fundamental skills and problem-solving aptitude. As we say, you can make a goat a better goat, not a tiger. Skilling only works when there is inherent talent. The National Education Policy’s Swayam Plus and the Academic Bank of Credits are promising, allowing credit-based, skill-focused learning that benefits deep tech readiness.
Q. How do you see contributing to national initiatives like ‘Make in India’ or ‘Atmanirbhar Bharat’, particularly in electronics and semiconductors?
A. Our core mission is to develop Atmanirbhar components. Deep-tech gives us technological sovereignty and flexibility. Take China’s civil-military fusion strategy—it’s shown how internally developed technologies can have broad strategic applications. India must develop technologies rooted in local problems. For instance, the West’s 5G infrastructure is designed for high-speed vehicles on autobahns. In India, our challenges are different—urban density, rural coverage, etc. At Pravartak, we are committed to creating India-specific technologies, including language models and secure systems, aligned with our national missions.
Q. You mentioned DST allows collaboration beyond IIT Madras. Could you share more about those partnerships?
A. Yes, we are not restricted to IIT Madras. The Department of Science and Technology (DST) has given us the flexibility to work with any Indian institution. We collaborate with institutes like IIITDM Kancheepuram, IIT Kanpur, IIT Mandi, IIT Hyderabad, and private institutions such as MIT ADT University in Pune. We even work with regional universities and remote colleges—for example, an arts college in Karur. We aim to discover and nurture talent across the country, not just in Tier-1 institutions.
Q. What initiatives have been taken to nurture STEM interest at the school level?
A. We launched Vidya Shakti during COVID to bridge rural-urban education gaps. Through CSR funding (from ICICI Bank, City Union Bank, etc.), we’ve set up smart classrooms with 65-inch TVs, local digital instructors (Digital Didis), and VR kits. The content is delivered in regional languages and supported by remote faculty. In one case, a student with no physical lab access scored 100 in Chemistry using only virtual labs. We’ve deployed 6,000+ centres in Andhra, 2,000+ in UP, 200 in Tamil Nadu, and even 15 in Sri Lanka.
Q. How do IIT Madras and DST support your innovation and partnerships?
A. IIT Madras is a foundational pillar for us. Around 60–70% of our solutions originate from its research ecosystem. We tap into startups incubated there, leverage faculty expertise, and access cutting-edge labs. It gives us a strong, credible pipeline for deep-tech problem-solving.
Q. What advice would you give to young entrepreneurs and researchers in electronics and embedded systems?
A. Focus not just on technology, but also on building a viable business. Listen to your customers—they often reveal insights that can reshape your product. Be open to change and never assume your first idea is the best one. Good products, constant feedback, adaptability, and wise spending are essential. Most importantly, stay humble and remember: in technology, resisting change is fatal.
Q. How do you integrate end-user feedback into R&D and design?
A. Pravartak operates as a bridge. We listen to the customer’s problem statements, understand what they need, and then consult with research institutions—primarily IIT Madras—to explore the best solutions. We don’t stop there. There’s a continuous feedback loop: the customer reviews, we take inputs back to researchers, refine the solution, and deliver an improved version. This iterative model ensures that deep-tech outcomes are both impactful and practical.
Q. Any final thoughts you’d like to share with India’s electronics and startup ecosystem?
A. Thank you for this opportunity. We’re deeply passionate about deep tech. Our journey wouldn’t be possible without IIT Madras’ leadership, especially the vision of our director and support from missions like the National Mission on Cyber Physical Systems. We are committed to ensuring that India not only innovates but also leads globally.
Q. Looking five years ahead, what is your vision for IITM Pravartak?
A. Personally, I aim to see at least 4–5 unicorns emerge from Pravartak-supported startups in the next five years. Beyond commercial success, I want these ventures to give back to the country. Just as the IT sector contributed immensely to India’s GDP in the 1990s, deep tech will define our future. Countries thrive on knowledge economies—India’s moment is now, and Pravartak aims to be a catalyst for that transformation.
