TechSparks 2026 grand finale: Intel, Jhunjhunwala, Food Pharmer on India’s rise

TechSparks is YourStory’s flagship startup-tech summit. It gathers entrepreneurs, investors, policymakers, and aspiring founders to exchange ideas and encourage innovation. The 17th edition focused on India’s ability to build rather than merely consume technology and products.

The grand finale explored how that shift could begin in classrooms, engineering colleges, and consumer markets. Speakers connected education, computing access, engineering practice, artificial intelligence, and trust in food labels. Their shared concern was structural: India has scale and talent, but its systems do not consistently turn them into products or opportunity.

The final day featured Sushma Boppana, Santhosh Viswanathan, Ashok Jhunjhunwala, and Revant Himatsingka, known as Food Pharmer. Their ideas ranged from personalised AI learning and hands-on engineering education to edge computing and transparent packaging. The closing message was direct: India’s next innovation chapter depends on people willing to take risks and build.

India’s education and employment challenge is enormous. The country has 247 million primary and secondary students to educate. At the same time, 250 million young Indians are expected to enter the workforce over the next five years. Preparing them requires more than increasing classroom capacity or test scores.

Sushma Boppana said resilience, engagement, and accountability matter more than academic education alone. Santhosh Viswanathan argued that technology should be built into everyday learning, because real work no longer separates technology from ordinary tasks. AI could personalise instruction around a student’s interests, making difficult subjects more accessible.

The current system remains rooted in nineteenth-century methods, even as society has entered the intelligence era. India also has limited computing access, which restricts technology-enabled learning. Viswanathan called the livelihood question both highly significant and potentially profitable. Better learning, stronger character, practical skills, and wider technology access will shape how these young people work.

Computing power is the ability of computers and data-centre systems to process information and run software. AI needs computing power to train models, handle data, and deliver answers or predictions. In practical terms, stronger access lets more people, companies, schools, and researchers use advanced digital tools.

Santhosh Viswanathan said India’s compute penetration is under 10%. He compared that with 95% in the United States and 75% in China. These figures describe a major access gap, not simply a difference in the number of talented engineers. A capable person still needs suitable hardware and infrastructure to develop or use demanding technologies.

The gap matters because 250 million young Indians are expected to enter the workforce within five years. India needs computing access to support learning, innovation, and livelihoods. Ashok Jhunjhunwala therefore encouraged a different AI strategy: compact, retrained open-weight models running on phones, laptops, and other low-power devices.

Computing power is a basic resource for modern AI development. Without enough of it, researchers and startups have fewer opportunities to train, test, and deploy systems. Learners and workers may also have less access to AI-enabled tools. The article links this shortage to India’s wider challenge of turning talent into technology and jobs.

Viswanathan reported that India’s compute penetration is under 10%, far below 95% in the United States and 75% in China. He also said 250 million young Indians will enter the workforce within five years. That combination makes access especially important: limited compute can restrict learning, entrepreneurship, and the creation of new products.

The article does not quantify the exact number of jobs or inventions lost. It does show the direction of the risk. India could remain a technology consumer instead of a builder. Jhunjhunwala’s edge-computing approach offers one response by using smaller models on phones, laptops, and low-power devices rather than relying only on enormous cloud systems.

India imports many technology products despite having capable engineers because the education and innovation system does not consistently teach people to build. Ashok Jhunjhunwala described this as a paradox: India can develop nearly every technology, yet imports nearly every product. The missing link is moving from technical knowledge to usable products.

He said the weakest part of engineering education is that students are not taught to build. His proposed mechanism is practical learning: discard half the current curriculum and use that space for hands-on creation. Students would gain experience turning ideas into working technologies rather than mainly completing academic requirements.

Jhunjhunwala also highlighted founders from Tier II, III, and IV cities as genuine risk-takers who can drive national transformation. Through ITEL, he is bringing scientists and startups together around civic problems such as urban transit and waste management. India’s talent can have greater impact when education, experimentation, and product-building are connected.

Edge computing means running computing and AI tasks on or near the device where they are needed, rather than sending everything to a distant cloud. Ashok Jhunjhunwala presented it as an alternative to chasing enormous, trillion-parameter models. The approach matters because India has a severe compute deficit, with penetration under 10%.

His proposal is to run compact, retrained open-weight models on phones, laptops, and low-power devices. Smaller models require less central infrastructure than huge cloud-based systems. Open-weight models can also be adapted for particular needs. Together, these features could bring useful AI closer to users and local problems.

Jhunjhunwala urged India to chart its own AI course instead of copying the largest-model race. The article does not claim that edge models solve every computing challenge. It presents them as a practical direction for a country with many potential users, limited compute access, and civic needs such as urban transit and waste management.

Food labels and factory access reduce information gaps by giving consumers clearer evidence before they buy. Companies know their ingredients, nutrition details, and production practices. Consumers often see only marketing. Showing specific information and opening operations makes the exchange more balanced and helps people judge products for themselves.

Revant Himatsingka’s brand, Only What’s Needed, puts its nutrition label on the front in large font. It shows every ingredient’s exact percentage in a pie chart and adds a QR code with labels in major Indian languages. OWN also invites people to factories and holds public board meetings. These mechanisms replace vague claims with accessible details and direct observation.

The brand has spent almost nothing on marketing. Its approach depends on transparency rather than conventional promotion. Himatsingka said that making a mistake could end the business, showing the risk of public accountability. The model suggests that clearer information and openness can help rebuild consumer trust, although the article does not measure its impact.