This week, Pixxel closed a $100 million Series C co-led by Temasek and Seraphim, the largest single funding round any Indian space technology company has ever raised. Six of its Firefly satellites are in orbit, reading the Earth in hundreds of bands of light. Its customers include NASA and the US National Reconnaissance Office. Its founders, Awais Ahmed and Kshitij Khandelwal, are barely thirty. The headlines will call it a triumph of Indian spacetech, and they will be right. But I have a smaller, more personal reason for writing about them, and it will take me half this piece to earn the right to tell it.
The story begins the way the best deep-tech stories seem to: with a problem nobody could buy their way around. As undergraduates at BITS Pilani in 2018, working on an AI challenge, Ahmed and Khandelwal wanted to predict the health of crops from satellite imagery, and discovered the imagery they needed did not exist at any price. Most students would have changed the project. They decided to change the imagery.
They had already worked together on the university's Hyperloop team; they understood hard engineering, and, perhaps more dangerously for two people in their early twenties, they were not intimidated by it. In 2019 they founded Pixxel to build hyperspectral satellites: instruments that read hundreds of narrow bands of light instead of a handful, and can see chemistry where ordinary cameras see colour, crop stress before the leaf turns, methane before anyone smells it, minerals beneath a hillside's surface.
"Most people adjust the dream when the first room says no. They did not. They adjusted the room."
Inference Editorial
The room that said yes early was in Los Angeles. In 2019, Pixxel was selected for the Techstars Starburst Space Accelerator, the only Asian startup in the batch, working alongside people from NASA JPL, the US Air Force and Lockheed Martin. Two students from Pilani, holding their own in a room full of the world's most established aerospace institutions. That was the real beginning. It was not, however, the first room they had asked to be let into. I know, because one of the earlier rooms was mine.
What followed was not a montage. It was seven years of the least glamorous work in technology, and it is the part of this story that deserves the most attention.
The Years Nobody Claps For
Consider what they were actually attempting. Two undergraduates, barely into their twenties, deciding to build satellites in a country where, at the time they started, spacecraft were something the government built. There was no Indian playbook for it, no queue of investors who understood hyperspectral imaging, and no shortage of sensible people explaining why students should not be doing this. They started anyway, before India's space reforms existed, betting that the rules would catch up with the ambition.
Then the world made it harder. They built their first spacecraft through the pandemic, with supply chains frozen and cleanrooms half-empty. Launch dates they did not control slipped again and again; a mission planned for 2020 did not fly until 2022. And in early 2021 came the moment that tells you who they are: in final testing, weeks before a scheduled launch, they found a software issue in their own satellite. They could have flown and hoped. Instead they pulled their spacecraft off the rocket, absorbed the delay and the doubt that came with it, and fixed the problem. In an industry where one bad day in orbit can end a young company, that decision, choosing embarrassment over risk, may be the most important launch they never made.
The demonstration satellites finally went up in 2022 and proved the thesis. The Firefly constellation, which Pixxel describes as the highest-resolution commercial hyperspectral system in the world, began deploying in 2025; six are in orbit today. Alongside the hardware came Aurora, the software that turns raw spectral data into answers, because the founders understood early that nobody buys pixels, people buy intelligence. None of it arrived on schedule. All of it arrived because they refused to stop.
Then the customers arrived, and they were not small ones. NASA. The US National Reconnaissance Office. Multiple iDEX wins with India's Ministry of Defence. Selection by IN-SPACe to lead India's first public-private Earth observation constellation of twelve satellites. And now the Series C: $100 million co-led by Temasek and Seraphim, with 360 ONE Asset, IMM Investment, Radical Ventures and growX participating, taking the company to $195 million raised. The money will fund the next constellation, called Honeybee, extending their sensing into new bands, alongside radar and ultra-high-resolution optical imaging. Pixxel now calls what it is building planetary infrastructure: a health monitor for the Earth itself.
Working on an AI challenge at BITS Pilani, Ahmed and Khandelwal discover the satellite imagery they need does not exist.
The company is born; Pixxel becomes the only Asian startup in the Techstars Starburst Space Accelerator, alongside NASA JPL, the US Air Force and Lockheed Martin.
Demonstration hyperspectral satellites reach orbit and prove the imaging thesis.
The commercial constellation begins launching; contracts follow from NASA, the NRO and India's Ministry of Defence.
$100M Series C co-led by Temasek and Seraphim; total funding reaches $195M; Honeybee constellation announced.
The Meeting That Stays With Me
A few years ago, when I was leading corporate innovation at T-Hub in Hyderabad, an email landed in my inbox from two young founders. They were building something bold in space technology, they wanted to explore T-Fund, and they asked if they could meet our CEO. I called them in right away and scheduled the meeting. It did not work out. Maybe we did not fully understand what they were building. Maybe T-Hub, and the ecosystem around us, simply was not ready for that kind of bet yet. I want to be careful here, because hindsight flatters everyone. I do not know that a different outcome in that meeting would have changed Pixxel's path, and I suspect the founders barely remember a conversation that we, on our side, have every reason to. They asked. We convened. The loop did not close. That is the whole record, and it is enough, because the sting is not about a missed deal. It is about what an ecosystem owes the people who show up early. Hyderabad could have been part of that first chapter. We had the conversation in the building. We just did not finish it.
The lesson is not that institutions should say yes to everything that sounds audacious. It is that the cost of a wrong no in deep tech is invisible for years, and then arrives all at once, denominated in someone else's headline. Ecosystems are not built by the capital they deploy but by the judgment calls they make when a technology does not yet fit any of their forms. Sometimes the most important thing an ecosystem can do is recognise the ones who refuse to wait for permission.
What I keep coming back to is the rarity of the thing itself. A clear, stubborn vision, plus the patience to keep going when the first room says no, is close to the rarest combination in entrepreneurship. Most people adjust the dream to fit the room. Ahmed and Khandelwal walked out and spent seven years building rooms of their own: in Los Angeles, in Bengaluru, in orbit. Rome was not built in a day. Neither, it turns out, is a health monitor for a planet.
So this is partly a correction of the record, published where I can see it every day. I am genuinely happy for both of them. They earned every bit of it, satellite by satellite, in the years when very few people were watching. Proud of what they are building for India and for the planet. Keep going. Onwards.
"The cost of a wrong no is invisible for years. Then it arrives all at once, denominated in someone else's headline."
Inference Editorial