There is a phrase that gets used loosely in startup circles. World-class from day one.

Most of the time, it is marketing. This time, it is not.

Sripurna Rao is 23 years old. He is from Surathkal, a small coastal town in Karnataka. He studied civil engineering at IIT Madras because he could not get aerospace with his JEE rank. And right now, working out of a 25,000-square-foot warehouse in Bengaluru with a team whose average age is 23, he is building an aircraft with an 8-meter wingspan that will fly at 45,000 feet for 24 hours continuously.

Not a drone. An aircraft.

The distinction matters to him enormously, and by the time you finish reading this story, it will matter to you too.

Arcturus Aerospace is building what could become the most cost-effective earth observation platform in the world. Where satellites charge enormous amounts for images taken once every 12 to 24 hours at one-meter resolution, Arcturus's aircraft flies continuously at high altitude, capturing centimeter-level resolution video for 95 percent less cost.

The investors who have backed this vision include Balaji Srinivasan, the former CTO of Coinbase and one of the most followed technology thinkers globally. And Srinivas Narayanan, the CTO of OpenAI. And South Park Commons, whose community includes ex-founders from Skydio and other leading aerospace and drone companies.

A 23-year-old from Surathkal. A $2.6 million pre-seed round. An aircraft that can see whether a person is holding a phone or a weapon from 45,000 feet.

This is his story.

For professionals tracking India's deep tech aerospace ecosystem, the AI Nexus World News Hub covers breakthrough Indian startup stories daily.

The Seventh Standard Spark That Started Everything

To understand Sripurna Rao, you need to go back to a seventh-grade science competition in Mangaluru.

His mother pushed him to participate in the National Children's Science Congress, a competition designed for 10th, 11th, and 12th standard students. He was in seventh standard. His team built something with biogas. They got selected at the district level, then the state level, then the national level.

The national-level event was held at IIT Varanasi. The top 20 teams got a month-long internship at IIT Kharagpur to explore laboratories and departments.

When Sripurna walked into the aerospace department at IIT Kharagpur, something clicked that has never unclicked.

They flew a model aircraft in front of him. They handed him the remote control. They let him feel what it meant to make something fly.

He found his first love.

By the time he returned to Mangaluru, he had told his father one thing. He absolutely had to build an airplane.

His father, a practical man, told him it was madness and it would pass in two days.

It has not passed. Sixteen years later, he is still building airplanes. Just significantly larger ones.

Why IIT Madras Civil Engineering Was the Best Decision He Never Made

When Sripurna wrote JEE in 2020, he got a rank of 4000. Aerospace was not available at any IIT with that rank.

He made a decision that he now calls the best decision of his life, even though it looked like a compromise at the time.

He looked at where three things were happening simultaneously: IIT, deep tech, and startups. The answer in 2020 was unambiguous. IIT Madras. Ather was breaking out. Agnikul was building rockets. ePlane was building electric VTOLs. GalaxEye was doing synthetic aperture radar imaging. All from IIT Madras.

He took civil engineering at IIT Madras and told his professors on day one: drone, drone, drone; airplane, airplane, airplane.

The professors were cooperative. The Centre for Innovation, a student-run facility open 24/7 with funding from alumni, became his home. The aero club became his laboratory. He became head of the product design club. He got patents.

And he started building.

The Path Through Crypto and Medical Delivery

Before Arcturus, Sripurna made his first startup attempt during COVID lockdowns.

Sitting at home with internet access installed for mock tests, he discovered Elon Musk's Falcon Heavy dual-landing video on YouTube. He spent six months reading everything about startups. Elon Musk. Bill Gates. The Indian ecosystem. From knowing zero to knowing everything in six months.

He noticed what was raging in 2021. Crypto.

He started with 5,000 rupees. Quickly realized trading was not the game. Identified a problem: ease of investment into DeFi for Indian users who could not make UPI payments directly into crypto. Built a platform in one month that let users pay via UPI, purchased crypto on their behalf, and gave them the keys.

The platform ran for a year. At its peak, 2 to 3 crore rupees of assets were invested through it. They got incubated in IIT Madras's student incubator. Bootstrapped completely.

Then the crypto crash of 2022-23 hit. Revenue disappeared overnight. They shut it down.

He learned everything that mattered. Logistics, regulations, paperwork, what a startup actually is, how it runs. Connections were built. When the time came for Arcturus, he already knew everything.

What Arcturus Aerospace Is Actually Building

Let Sripurna explain it in his own terms, as he did on the Mundhe Banni podcast.

There are currently two main options for earth observation from above. Satellites in orbit and drones close to the ground.

Satellites cost enormous amounts to launch. They take images at 1-meter resolution every 12 to 24 hours as they orbit. The best military satellites reach 0.5-meter resolution but are restricted from civilian use because at that resolution you can see military installations clearly.

Drones fly for 15 to 20 minutes maximum. To cover India adequately for any commercial earth observation use case, you would need 10,000 drones. The infrastructure cost is impossible.

In the middle, flying at 40,000 to 45,000 feet for 24 continuous hours, is where Arcturus sits. High-altitude long-endurance aircraft. There are only four or five companies in the world that build at this altitude and endurance level. Every single one of them builds for the military.

No commercial option exists in this category. So the market above them has defaulted to satellites. Not because satellites are the best option. Because the commercial aircraft option does not exist.

Until Arcturus.

Here is what makes Arcturus's aircraft different from those four or five military players:

The technology is 20 to 25 years old in those companies. Materials science, electronics, avionics, and telecommunications have all changed dramatically. Arcturus is building with current technology, achieving the same capabilities at dramatically lower operational cost.

They control the entire stack. From airframe structures to avionics, from electronics to software, from telecommunications to the optical payload integration. Every component they can build, they build. For components they cannot yet build, they import sub-components and assemble them in-house. Controlling the stack means they can design for the specific end use case rather than adapting a general military platform.

The economics are radically different. Current satellites provide images at commercial pricing that makes the data "good to have" but not operationally essential. Arcturus is targeting operational essentiality: continuous 24-hour time-lapse video at centimeter resolution that changes how oil and gas refineries monitor contractors, how insurance companies validate claims, how wildlife authorities track animal movements, and how agricultural operations optimize inputs.

The resolution is the game changer. From 45,000 feet, Arcturus's aircraft captures images at centimeter-level resolution. At this resolution, you can see whether a person is holding a phone or a weapon. You can see whether a car has a petrol can or a bomb in its back seat. You can see whether contractors at a refinery are following safety protocols or not.

The aircraft currently flying proof-of-concept missions has an 8-meter wingspan, a 4-meter body length, can carry 50 to 250 kilograms of payload, and flies for 10 to 24 hours. The final product will reach 45,000 feet, carry 250 kilograms of payload including multiple camera types simultaneously, and fly for 24 hours continuously.

The Commercial Use Cases That Make This a Billion-Dollar Opportunity

Sripurna's insight about the addressable market is precise and worth understanding in detail.

The commercial customers who currently buy satellite earth observation data are concentrated in specific industries: agriculture, insurance, oil and gas, shipping, and logistics. These are the five biggest commercial plays.

Here is the critical observation: all of these industries operate at fixed points. Refineries do not move. Pipelines stay in place for 10 years after a capital expenditure is committed. Agricultural land does not shift overnight. Insurance claim locations are specific and documented.

If 99 percent of Earth's surface is covered by satellites taking general images, but the commercial customers who derive real value from Earth observation are concentrated at approximately 1 percent of the Earth's surface at fixed, known locations, then the economics of the situation become clear.

You do not need to cover the entire Earth continuously. You need to park an aircraft above the specific fixed-point industrial installations that will pay for continuous high-resolution monitoring.

At centimeter resolution with 24-hour continuous coverage, the value proposition to each of these industries is fundamentally different from what satellites can currently provide.

Oil and gas: A refinery like MRPL in Mangaluru has contractors, pipelines, equipment, and loading operations. Currently, satellite data provides validation of contractor-provided information, not a primary data source. At centimeter resolution with continuous coverage, whether contractors are following SOPs, whether equipment is being properly utilized, or whether reported incidents actually occurred as described: all of this becomes directly verifiable from a single aerial platform.

Agriculture: Weed patterns from above with daily imaging reveal exactly where pesticides need to be applied. Soil health through hyperspectral imaging shows nitrogen and potassium levels field by field. Water flow patterns show where fertilizers leach. A farm operator can optimize inputs to a precision that was previously impossible at scale.

Insurance: Instead of applying district-level historical rainfall averages to set premiums, insurance companies can use specific location data about actual conditions over six months to price risk accurately. Policyholders who have experienced genuine losses can be validated faster and with more precision.

Wildlife management: WWF is already in conversations with Arcturus. The ability to monitor large wildlife areas through infrared imaging for 10 hours continuously at night changes wildlife management from reactive to predictive. Animals have patterns. Enough data makes those patterns predictable.

The Veerappan point: As the podcast hosts noted, if this technology had existed 30 years ago, locating Veerappan after the Rajkumar kidnapping would have taken a day rather than 108 days. The forest is vast, but it is fixed. An aircraft at 45,000 feet with infrared and synthetic aperture radar can see through the forest canopy.

The Funding Journey: From 20,000 Rupees in the Bank to Balaji Srinivasan's DMs

The funding story is worth telling in full because it illustrates something important about how India's deep tech ecosystem is changing.

Sripurna graduated from IIT Madras with approximately 15,000 to 20,000 rupees in his bank account. He had turned down placement interviews by flying home to Mangaluru the day before Day One so his professors and batchmates could not pressure him to attend.

He moved into a PG in Electronic City, Bengaluru, near a warehouse belonging to Airbound, another Grad Capital portfolio company. His intern was his roommate. They rented Yulu bikes daily to commute because at 2 rupees per minute for a 10-minute journey, 20 rupees was the optimum. He timed his rides to save 4 rupees by reaching them in 8 minutes.

Grad Capital had given him $5,000. He spent it the next day: the 3D printer, materials, and motors were all ordered immediately. Within a month, he had built a proof of concept that could fly from Bengaluru to Chennai autonomously, covering 300-350 kilometers continuously.

They invested $40,000. The round was beginning.

One night in Bengaluru, his PG had no room. He took a bus to Majestic for 10 rupees with his bag. He slept at the bus stand rather than call home for money. He had decided not to ask for money from home and stuck to it through sheer stubbornness.

From there, things began to move.

He posted a tweet about what he was building. He was not particularly active on Twitter. The tweet went viral because US-based investors reposted it. Then Balaji Srinivasan slid into his DMs: "If you are building a company, send me an email."

He checked twice that it was really Balaji and not a fake account. Then he emailed immediately. The reply came back fast. Balaji said he wanted to invest.

South Park Commons invested after Sripurna spent a month in their community, where ex-founders from Skydio and other leading aerospace companies validated what he was building.

Version One's Boris, who built and sold a company to Amazon before opening his own fund, invested.

The IIT Madras alumni connection brought Srinivas Narayanan, OpenAI's CTO and fellow IIT Madras alumnus, into the conversation through South Park Commons. The aerial data angle was directly relevant: OpenAI needs training data for multimodal models, and high-clarity aerial data at the resolution Arcturus captures does not exist elsewhere at scale.

Total raised in pre-seed: $2.6 million.

The Team, the Warehouse, and the Physics

Sripurna's team of 10 engineers has an average age of 23. He specifically sought engineers who came from IIT Madras's student project culture: the rocketry team, the autonomous car team, and the Aero Club. People who had learned to build complex systems from scratch without textbook guidance.

The warehouse is 25,000 square feet in the Hennur area of Bengaluru. Six months ago they were assembling aircraft in two bedrooms of a 3BHK apartment in Bhartiya City. Inventory in two rooms, assembly in the hall.

The aircraft they are currently building will fly at 20,000 feet and be ready to fly within a month of this interview. The final product targeting 45,000 feet with 24-hour endurance is planned for year-end 2026.

The engineering challenge they have to solve on the way from the ground to the stratosphere is genuinely complex. Temperature goes from 24 to 34 degrees Celsius at ground level to minus 25 to 30 degrees at 40,000 feet. Pressure drops exponentially. Air density changes continuously. Wind speeds and conditions shift throughout the ascent and at the operational altitude.

Every subsystem must be designed to function through all of these conditions simultaneously. Not just at one altitude. Through the entire flight envelope from takeoff to maximum altitude and back.

This is why Sripurna says aerospace is as much art as engineering. The documentation is publicly available for any aircraft ever built. But the understanding of how complex systems interact, the rules of thumb built through decades of failure and iteration, and the judgment about manufacturing tolerances that only come through hands-on experience: these cannot be acquired from papers alone.

His response to this gap was to build a network of mentors. Through his investors' portfolio connections, he talks regularly with engineers at Boom Supersonic and Hermeus in the US, companies that have been building complex aerospace systems for years. What took those companies 50 years of accumulated experience, he is absorbing in six months of conversations.

The Business Model: Data, Not Hardware

Arcturus is not an aircraft company that sells aircraft. It is a data company that collects data.

The business model is closer to Zepto than to Boeing. Zepto identified that dark stores placed strategically could serve maximum customers at minimum logistics cost. Arcturus positions airstrips strategically to serve maximum industrial customers at minimum operational cost per hour.

Customers pay for images or video clips, not for flight hours. Arcturus optimizes operations to serve multiple customers from each flight, amortizing operational costs across several revenue streams simultaneously.

Current unit economics from early validation: customers in early conversations were willing to pay 10,000 rupees per hour. Arcturus's operational cost at that stage was 500 rupees per hour. The market rate for satellite alternatives was 50,000 rupees. At 10,000 per hour from a customer who sees the satellite alternative at 50,000 and an operational cost of 500, the margin structure is extraordinary.

The final product with a 250-kilogram payload capacity can carry five different camera types simultaneously: optical, infrared, hyperspectral, and synthetic aperture radar. Every customer type gets the data they need in a single flight. The operational efficiency improves with every customer added to the route.

Privacy, Ethics, and the Right Use Cases

The podcast hosts raised the question that any honest analysis of this technology must address. A platform that can see whether someone is holding a phone or a weapon from 45,000 feet raises legitimate privacy concerns.

Sripurna's answer is straightforward and worth examining honestly.

The commercial economics of industrial monitoring make human surveillance a poor business choice for Arcturus. A refinery paying for continuous monitoring generates 100 times more revenue than any human tracking application.

The military already has this capability. Any government that wants to conduct aerial surveillance has the platforms to do it. Arcturus entering or not entering this market does not change what governments can do.

The regulatory framework, both in India through DGCA and internationally through FAA-derived standards that India largely follows, requires transponders that broadcast aircraft position to air traffic control. Operations require a licensed pilot monitoring from the ground even for autonomous flights.

These are not complete answers to privacy concerns around a technology this capable. But they are an honest framing of how the technology is actually deployed commercially and what limits exist.

What This Story Means for India's Deep Tech Ecosystem

Sripurna's story is not just inspiring at the individual level. It is a signal about something structural that is changing in India.

Pixxel, started by BITS Pilani students, now operates hyperspectral satellites and sells data to global customers from Bengaluru. Agnikul, from IIT Madras, is building rockets. Ather, from IIT Madras, is producing what many consider the world's best electric scooters.

Each of these companies created a proof of concept for the next one. They showed that Indian students can build things of genuine global quality. They built the investor ecosystem, the mentor network, and the talent pipeline. They reduced the time to credibility for the generation that followed them.

Ather took 10 years. Pixxel took five. Sripurna believes Arcturus can do it in two to three, building on everything those companies established.

The Karnataka government's Elevate grant program paid out within four days of the award. Bengaluru's startup ecosystem operates with a high-trust abundance mindset that enables this. South Park Commons opened an India office in Bengaluru. The alumni network of IIT Madras provides capital, mentorship, and introductions that were not available to founders a decade ago.

India is building deep tech infrastructure. Slowly. Imperfectly. But unmistakably so.

For professionals tracking India's AI and deep tech ecosystem alongside global developments, the AI Nexus World Community connects builders, investors, and professionals navigating this landscape.

Advice for Students Who Want to Build

When asked what students interested in aerospace or any deep tech field should do, Sripurna's answer was characteristically direct.

All the information is available online. He is a self-taught aerospace engineer. Every technical book he needed was available as a PDF. The literature for building any aircraft, including supersonic fighter jets, is publicly available.

The challenge is not access to information. The challenge is starting.

Most people believe that aerospace is too complicated, that it requires specialized resources they do not have, and that it is reserved for people with the right backgrounds and institutional support.

Sripurna built his first aircraft in Mangaluru from components he bought at Kinetic Hobbies near PVS Circle. He learned to fly it from an online simulator he found and installed himself. He learned from Flite Test, an online resource he accessed in seventh and eighth grade.

The starting point was not an IIT lab or an industry internship. It was curiosity, a local hobby shop, and the internet that led me to it.

From there to 45,000 feet is a long journey. But it starts with starting.

Citations

Mundhe Banni Podcast, Sripurna Rao Episode https://www.youtube.com
IIT Madras Centre for Innovation https://www.iitm.ac.in
Grad Capital Official https://www.gradcapital.in
South Park Commons https://www.southparkcommons.com
Karnataka Elevate Programme https://elevate.karnataka.gov.in
NASSCOM, India Deep Tech Report: https://nasscom.in