Srinath Ravichandran, Co-founder & CEO of spacetech startup Agnikul Cosmos, speaks to BW Businessworld about company's “Mission 2,” rocket reusability, orbital data centres and why launch bottlenecks are reshaping the global space market. Excerpts:
Where does Agnikul stand today, and what does 2026 represent for the company?
We’re building a rocket to take satellites to space on demand, and we’re doing it very differently from the rest of the market. Most rockets today are one-time-use, and we don’t think that’s a sustainable business model. We’re building completely reusable systems.
Small rockets, historically, were never optimised for production or profitability. They were often seen as demonstrators, a stepping stone to something bigger. We had to redesign everything from scratch to make a small rocket commercially viable. That’s what led us to technologies like 3D printing, engine clustering and a mobile launch pedestal.
In 2024, we demonstrated these technologies in a fully successful controlled suborbital launch. It wasn’t a sounding rocket, it was a guided flight, steered through winds and launch-day uncertainties to a precise point in space. Mission 2 is the next step.
What are the key objectives of Mission 2?
Mission 2 has two objectives. The primary one is commercial, which is taking customers to orbit. The secondary objective, which is critical for long-term profitability, is recovery and reuse.
We will attempt to bring back the lower stage, while the upper stage continues in orbit and performs work for us. Recovery is a test objective because this is the first time we’re attempting it, and no one in India has recovered a rocket stage in this form before. Extending the upper stage’s life to this level is also new for the country.
You’ve spoken a lot about reusability. Why is it so central to your strategy?
Reuse has been built into the vehicle from day one. That’s one of the smartest strategic decisions SpaceX made and they didn’t treat reusability as an afterthought.
Every time you fly the same hardware again, you learn. Vehicles don’t always behave exactly the way models predict. If you throw hardware away after every flight, you lose that learning. Our ascent trajectory is something we’ve already demonstrated and have a good handle on. Recovery is new, but we didn’t want to fly another mission without attempting it.
How do orbital data centres fit into Agnikul’s business model?
Because our upper stage stays active after deploying a satellite, it effectively becomes a platform. It can host data centres, communication transponders, cameras or hyperspectral imaging payloads.
The key idea is that the upper stage itself becomes the satellite platform. A data-centre operator doesn’t need to build a separate satellite — we host it on the rocket stage. That makes the system lighter, less expensive and cheaper to launch. While the upper stage is in orbit, it generates revenue for us.
What are the technical challenges in making orbital data centres viable?
The biggest bottleneck is the GPU stack. From our side, navigation, telemetry, power generation and radiative cooling are all solvable problems. GPUs, however, need to be radiation-tolerant and space-hardened.
Right now, there’s no equivalent of Nvidia for space-grade GPUs. That’s the real gap. Once that gets solved, a lot of these applications become fully commercial. Until then, it’s a mix of experimentation and early deployments.
How strong is commercial demand right now for small-satellite launches?
Demand is very real. Customers gained confidence after our previous flight. At the same time, people are feeling the Falcon 9 shortage much more acutely now.
SpaceX is focusing more on Starship, which means fewer Falcon missions. Small satellite operators are often pushed into ride-share missions through aggregators, who charge a premium and don’t offer schedule certainty. That’s where a dedicated, flexible launch option becomes attractive, especially when recovery starts improving the economics.
Who are your core customers globally?
There are three broad categories. First, the large communication and imaging companies you’d recognise globally. Second, very early-stage startups in India, Europe and Australia.
But our core market is the middle group, companies that are already commercial, making real money, and for whom launch is a critical supply-chain input. These firms are severely deprived of reliable launch options today, particularly in Europe and Australia. They need fixed schedules, and that’s what’s missing in the market.
What kind of applications dominate this demand?
Imaging and communication together account for about 80 per cent of the market, with mass to orbit split roughly evenly between the two.
We’re also seeing interest from companies that want to test subsystems in orbit to gain space heritage. That’s important because large satellite builders look for proven, flight-tested components. Space manufacturing and data centres are smaller today, but there’s meaningful capital flowing into those areas.
How do Indian customers compare with international ones?
India is interesting because most satellite startups here emerged only after policy reforms through IN-SPACe, which means many are still early-stage.
We’re speaking to all of them. The community is small, and we know many founders personally. While we don’t have signed contracts yet, the conversion pipeline for Indian customers is shorter because they know us, they visit our facilities in Chennai, and they understand what we’re building.
How do you see competition evolving globally in small launch vehicles?
Today, the only major suppliers are Falcon 9 and Rocket Lab’s Electron. Many countries are trying to build launch capability, but most are still in the proving phase.
In the long run, say 10 years, I think only five or six small launch companies globally will represent the steady state. This is extremely difficult engineering, and it’s a tough business. There will be consolidation, and geopolitics will influence who customers choose.
Where do you see India’s space opportunity heading over the next two years?
The biggest immediate bottleneck remains launch, and it’s only getting more severe. In 2026, the focus will be on getting core payloads right — things like synthetic aperture radar and optical sensors — and on improving launch availability. More unconventional applications like space manufacturing and operational orbital data centres will start becoming visible around 2027. But for now, it’s about solving the fundamentals. |
