Pixxel's Awais Ahmed: Govt Contracts Help Spacetech Start-Ups Establish A Strong Base Early On

It's very important. When we began in 2019, there was no semblance of a private space ecosystem. So, it was not possible to get any kind of contracts, let alone pre-contracts.

Once we demonstrated some capability on our own, the contracts came in. We had already launched and demonstrated our technology, and then we participated in the entire bid process to win contracts on the basis of our merits.

Government contracts help with fundraising, cash flow and establishing a stronger base for a company early on.

In the US, a company that starts today with a good team and some early science can get nearly $20mn of grant money for building the technology that the government would need, and then an automatic contract.

Once you develop it, it can go all the way up to $300mn as a contract. This is only in the first two to three years of a company's existence. When you have access to that kind of revenue, it makes it easier for investors to come in.

It is not a fair comparison because the US has been doing this for decades. India is doing it this decade, but the government is rapidly starting to give contracts.

When we started, people only invested in software-as-a-service and fintech—not even in hardware, let alone space hardware. Now it's gotten much easier, especially in the early stages. Even the government is discussing funds for deep-tech investment. In the early stages, the challenge was to prove that this was a viable business, that if we build a satellite and deliver data, there is a market, and that the team can deliver. Most of the early investment came on the basis of the team's strength.

We had built the Hyperloop pod [a capsule designed to carry people or cargo through low-pressure tubes] in India, taken it to the SpaceX headquarters and presented it to Elon Musk. Building a Hyperloop pod was a very challenging technical problem for a bunch of students without access to a lot of money. The fact that we had managed to put that together, take it there and meet him gave investors confidence that this was a team probably worth backing.

Before that, I was part of the student satellite team at Birla Institute of Technology and Science, Pilani, where we were working with scientists at Isro [Indian Space Research Organisation].

When we raised our Series A, there were early signs of customer interest and contracts. We had one satellite up in space. We had proven that we could put it into orbit, that it would work and beam down data, and that the entire technical chain had been validated.

The later round was around the commercial strength of the company's order book—how many customers we have, how global it is and how large these contracts could get. So, it was the strength of the team and the idea, then proof of the technology and finally the commercial order book.

I can't speak of the numbers, but I can give you a sense of some customers. We are working with the Indian Air Force, [US space agency] Nasa and the National Reconnaissance Office in the US. Nasa buys data from our hyperspectral satellites.

We also have contracts with Rio Tinto [a global mining company] and a bunch of others. The clientele is global, although it is concentrated in the US and India. We also have customers in Europe, West Asia and South America. It includes both government and private clients, across data and satellites. We are close to 100 customers globally. About 90% may be commercial and the rest are government organisations in in India and the US.

Private-sector customers are predominantly in agriculture, oil and gas, and mining.

We launched a constellation of six Fireflies [the world’s highest-resolution commercial-grade hyperspectral satellites] last year. We are launching at least 12 more satellites next year and early 2028, and it will expand further.

We are building 12 hyperspectral satellites, which we will own and sell the imagery from, including to Indian government users. We also have contracts where satellites are being built, delivered and operated for the government. The IAF contract involves a 200kg modular satellite for hosting four different kinds of payloads.

We design, integrate and test the satellite 100% in India. But no one does all of it indigenously because some parts are not available. For example, imaging detectors and high-efficiency solar cells have to

be imported. Everything else—from the structure, thermal aspects and onboard computer to electronics, avionics, software and mission control—is built locally. So, 60-65% of the satellite is being built locally, with 35-40% components not available here.

There are companies working on it. Once you have a certain level of foundry capability, you can start doing detectors as well. Solar cells are also being developed, although they require quality at the silicon-cell level. 

A lot of that capability is being built. Pixxel itself is bringing a lot of sensors and actuators in-house that would otherwise have been imported. Probably, fast forward seven to 10 years, that 60% will end up being 90%.

There are two aspects: technology and cost. The economics of large-scale data centres in orbit have yet to be proved. Likely, the only company in the world that could make that work in the short term is SpaceX because they don't have to pay margins on their launch. 

But the issue is that it will get tougher to build data centres on Earth, because approvals will get harder and timelines longer. You don't want to be polluting neighbourhoods and areas. So, even if you spend more money in setting up data centres in space, you are able to start getting more compute.

If US companies are utilising this for various purposes, including defence and intelligence, India needs to have that capability. It was important for us to say that this is a problem that can be technologically solved. 

All you have to do is put GPUs [graphics processing units] in orbit and make sure that they are normally managed. Who better than Sarvam [artificial intelligence start-up] to partner with in India, given the indigenous models and their training. We are good at building satellites and putting hardware in orbit, while they come in with the AI workloads and training. 

There are technologies that you need to build as a country regardless of the economics.

Correct. You can get more power through the Sun, and the cooling aspect is also quite natural there. But you can't just let the chips be because vacuum doesn't conduct heat; you have to radiate it away. 

There is a lot of thermal engineering to be done, including a thermal loop and radiators that radiate heat into the vacuum. Power-wise, you have unlimited coverage if you put them in the right orbit because solar panels are always facing the Sun.

We started launching satellites in January and August, and then started selling data a few months after the launches. In the 2025 calendar year, we became fully commercial. 

Last year, we also became fully commercial in our satellite business, where we build and provide satellites to various government and private organisations. So, both data, which is our planetary intelligence business, and our satellite business, our planetary infrastructure business, properly got commercialised last year. 

We did revenues and sold to customers even before that, but it reached proper commercial scale in 2025.

No, the investment was purely financial. Separately from the investment, we are in conversations for a variety of strategic things. 

They thought it was a good company and saw the potential future and said they would like to invest. It's a good backing to have, given the kind of capital they have. 

Separately, we continue to see what we can work on together, whether it's Google Earth or something else, but the investment was purely financial.

It was a $24mn Series B in December 2024. We had raised $36mn a few weeks before that. So, the total is $60mn as part of the same round. Overall, we have raised $96mn. 

At this point, we are more focused on execution and serving our customers. We have the capital to execute most of our plans.