TechSparks 2026 Day 1: Sridhar Vembu, Priyank Kharge, spacetech founders chart India's 'New Tech Order'

TechSparks is YourStory’s flagship startup-tech summit. Its 17th edition opened in Bengaluru with founders, investors, policymakers and technologists. The event examined how India can build a stronger technology ecosystem. Its central message was that prosperity cannot simply be imported through foreign products, capital or intellectual property.

The “New Tech Order” theme focused on moving beyond technology services. Speakers discussed infrastructure, foundational technologies, artificial intelligence, spacetech and regional innovation. Sridhar Vembu argued that ownership matters more than simply performing technical work for global companies. Priyank Kharge described efforts to spread investment beyond Bengaluru. AgniKul Cosmos and GalaxEye shared lessons from building private space businesses.

The summit also launched Tech30 2026, a selection of 30 promising startups. The group spans spacetech, defence, clean energy, advanced materials, agritech and healthcare. Together, the sessions presented a clear direction: India should develop and own technologies that solve its problems, while creating value that remains in the country.

Providing technology services means doing development, design or support work for companies that own the underlying products and intellectual property. Owning foundational technology means creating the essential systems, hardware, software or platforms that other products depend on. The difference matters because ownership controls who captures the largest share of economic value.

Sridhar Vembu gave a clear example. He claimed India performs 40–50% of Nvidia’s chip design work and much software development for Microsoft, Oracle and Cisco. Yet foreign companies hold the relevant intellectual property, so financial returns flow mainly to those owners. India supplies skilled work but does not necessarily control the technology or its profits.

The proposed shift would require investment in infrastructure and deeptech hardware, not only services. Vembu said Zoho is redeploying engineers to build server infrastructure in-house. The broader goal is a technology economy where Indian companies own foundational layers and generate more prosperity for non-tech workers too.

Karnataka’s technology base outside Bengaluru is substantial. The figures cited at TechSparks total close to Rs 10,500 crore in IT exports across Mysuru, Mangaluru and the Hubballi-Belagavi-Dharwad cluster. These numbers challenge the idea that Bengaluru is the state’s only meaningful technology centre.

Mysuru delivers close to Rs 4,000 crore in IT exports. Mangaluru contributes close to Rs 3,500 crore. The Hubballi-Belagavi-Dharwad cluster contributes close to Rs 3,000 crore. Added together, the three figures come to approximately Rs 10,500 crore, although the article describes each number as “close to.”

Priyank Kharge presented these results as evidence for Karnataka’s Beyond Bengaluru Mission. The policy aims to take investment to places where talent already lives and reduce pressure on Bengaluru. Region-specific centres of excellence are also planned or being established, including AI centres in Dharwad and Raichur, an agritech centre in Kalaburagi and a horticulture centre in Bagalkot.

Sridhar Vembu argued that India should strengthen its infrastructure and foundational technology before making massive GPU purchases. GPUs support AI computing, but buying them does not automatically give India control over the systems, products or intellectual property built with that computing power. His concern is that spending can create capacity without creating lasting ownership.

He illustrated the point by saying that, as a government, he would spend $10 billion on infrastructure in Bengaluru, Chennai and elsewhere rather than on GPUs “right now.” He also claimed India performs 40–50% of Nvidia’s chip design work, while foreign companies retain ownership of the resulting technology and receive the financial returns.

Vembu therefore places infrastructure and foundational layers first. He said India should learn from East Asian economies that built core capabilities before reaching the next stage. The implication is that India needs domestic technology ownership, reliable infrastructure and deeptech capacity before expensive AI hardware can deliver broad national prosperity.

AI-written software could sharply reduce the amount of routine coding done by human engineers. Vembu compared the change to machine looms disrupting handloom work. His estimate was that AI already writes better code than 60–70% of software engineers and could outperform 90% within three to six months. If that happens, coding roles would likely shift toward higher-value technical work.

Zoho’s response is to redeploy engineers rather than lay them off. Vembu said the company is moving people into deeptech hardware, including building server infrastructure in-house. This uses existing engineering talent for work that may require hardware knowledge, systems thinking and control over foundational technology.

The forecast describes a fast transition, but the article presents it as Vembu’s estimate rather than a measured industry result. Its broader implication is clear: software companies may need to retrain workers and build deeper technical capabilities. Engineers could spend less time writing routine code and more time creating infrastructure and hardware.

AgniKul Cosmos and GalaxEye matter because they are building private-sector capabilities in space. AgniKul is pursuing commercial rocket launches, while GalaxEye is developing satellite-based Earth observation. Their work expands the industry beyond established public institutions and targets markets with commercial and national importance.

Their experiences show the sector’s difficult operating model. AgniKul faced 275 investor rejections, and GalaxEye faced around 100. AgniKul Co-founder Srinath Ravichandran described a frugal pay-per-kilogram approach, aiming for commercial viability at $50–60 million rather than depending on enormous funding. GalaxEye Co-founder Suyash Singh recounted losing communication with its satellite 30 days after launch.

These setbacks reveal that private space companies must survive long development cycles, technical failures and cautious investors. The founders treated rejection and failure as part of building the industry. Singh also identified defence as one of the largest markets for space-based Earth observation, making geopolitical demand a significant future driver.

Intellectual property, or IP, is the legal ownership of creations such as inventions, software, designs, technical methods and brands. It allows the owner to control how the creation is used and to earn from licences, products or services built around it. Foundational technology is especially valuable because many other products can depend on it.

Vembu argued that India performs substantial technical work but often does not own the resulting IP. He claimed Indian teams handle 40–50% of Nvidia’s chip design work and large amounts of software development for Microsoft, Oracle and Cisco. In his view, the foreign owners capture the main financial returns because they control the underlying technology.

That ownership affects more than company profits. Vembu said it determines the income level of non-tech people in an economy. If Indian firms own foundational technologies, more value can remain in India and support broader prosperity. If they remain service providers, growth may depend mainly on fees paid by foreign IP holders.