The startup that built India's most ambitious private satellite — and then lost it

In 2019, a bunch of students from IIT Madras flew to California to compete in Elon Musk's Hyperloop competition. They didn't win. But they came back with something more useful — the conviction that a small Indian team could build hard, physical, deep-tech things.

Two years later, five of them — Suyash Singh, Denil Chawda, Kishan Thakkar, Pranit Mehta and Rakshit Bhatt — started a company called GalaxEye.

And they picked one of the hardest problems in space to solve.

Here's the thing about looking at Earth from orbit. You basically have two options.

Option one is an optical camera. It takes photographs, essentially. Beautiful, colour-rich, instantly understandable images. Anyone can look at one and figure out what's happening. The problem? A camera is only as good as the light. Cloud cover? Blind. Night time? Blind. Monsoon? Forget it. And India happens to spend a large chunk of the year under cloud.

Option two is Synthetic Aperture Radar, or SAR. Instead of collecting light, SAR fires microwave pulses at the ground and reads what bounces back. Clouds don't matter. Darkness doesn't matter. It works 24x7 in any weather. The catch is that SAR images look like grainy black-and-white noise to anyone who isn't a trained analyst. It's data that needs a specialist standing between it and the customer.

So the industry made you choose. Pretty but unreliable, or reliable but unreadable.

GalaxEye's pitch was: why not both? On the same satellite. Pointed at the same patch of ground. At the same instant.

They called it OptoSAR — an optical sensor and a radar sensor mounted on a single platform, with the two data streams captured in one pass and stitched together by onboard AI. GalaxEye branded the fusion technique "SyncFused" and claimed it delivers up to three times more actionable information than a single-sensor satellite. They even put an NVIDIA Jetson Orin computer on board so a lot of the processing happens in orbit rather than on the ground, cutting the time between "satellite sees something" and "customer gets an answer."

The business model was Data-as-a-Service. Don't sell satellites. Sell subscriptions to what the satellites see — to defence, agriculture, insurance, infrastructure, shipping, disaster response. Then scale it into a constellation of 8-12 satellites by the end of the decade, so you can revisit the same spot every few hours instead of every few days.

For a country that imports a great deal of its commercial satellite imagery, this was a genuinely attractive idea. Investors thought so too. Infosys, Speciale Invest, Mela Ventures, Rainmatter, ideaForge and MountTech Growth Fund — a defence-focused fund set up by former Defence Secretary Ajay Kumar — all wrote cheques. In February 2026, NewSpace India Limited, ISRO's commercial arm, signed on to resell GalaxEye's imagery.

And on 3 May 2026, it finally flew.

Mission Drishti lifted off from Vandenberg Space Force Base on a SpaceX Falcon 9. At roughly 190 kg, it was the largest Earth observation satellite ever built by a private Indian company, and the world's first commercial satellite carrying both radar and optical sensors together. The Prime Minister tweeted his congratulations. The Indian Space Association called it a pivotal moment. Five years of R&D, vindicated in about nine minutes of flight.

Then space did what space does.

During the satellite's Launch and Early Orbit Phase — the most delicate stretch of any mission, when a spacecraft is still waking up and getting its bearings — a severe geomagnetic solar storm hit.

Solar storms aren't gentle. When the Sun throws off a large burst of charged particles and it slams into Earth's magnetosphere, the radiation environment in low Earth orbit spikes hard. Satellites are built with shielding and redundancy for exactly this. But "built for it" and "survives it" are different sentences, especially for a first-of-its-kind spacecraft that has never been stress-tested by reality.

GalaxEye's initial root-cause analysis pointed to radiation from the storm damaging a critical onboard subsystem. Communication with Drishti first became patchy. Then it stopped. Independent satellite trackers had reportedly spotted the spacecraft tumbling in orbit through May, before the company publicly confirmed anything.

On 7 July 2026, roughly two months after launch, GalaxEye said the odds of recovering the satellite were low.

Now, it's worth being precise about what failed and what didn't. The rocket worked. The deployment worked. The satellite powered up, talked to GalaxEye's own Mission Control Centre in Bengaluru, and validated several subsystems over its first few weeks in orbit. This wasn't a design flaw that killed it on day one. It was an environmental event that killed it on roughly day forty.

But here's the uncomfortable part. In the Earth observation business, you don't get paid for building satellites. You get paid for the data. And GalaxEye never got to the point of proving that OptoSAR — the entire thesis of the company, the thing five years and millions of dollars went into — actually delivers fused, commercially useful imagery from orbit.

The technology remains unproven. Not disproven. Unproven. Which, when you're going back to investors and defence buyers, is an awkward place to stand.

Here's where it gets interesting.

If you pull GalaxEye's share allotment filings, the company has raised roughly ₹212 crore since inception — about $24 million across five years. For a company attempting to build something no one in the world had built before, that is a remarkably thin cheque book. ICEYE, the Finnish SAR company GalaxEye is often compared to, has raised over $500 million.

Look at how the price per share has moved:

Almost a 15x on paper in under five years. But notice the shape of it. The big leap happened early — nearly 6x between 2021 and 2022, when GalaxEye had little more than a prototype and a pitch. Since then the step-ups have been getting flatter: 1.9x, then 1.3x. That's what an investor base does when it's still willing to fund you, but wants to see hardware in orbit before it re-rates you properly.

And then there's the timing.

GalaxEye's single largest allotment ever — ₹49.29 crore, on 10 June 2026 — landed five weeks after Drishti launched. Add the three allotments in January 2026 and the company pulled in about ₹93 crore during 2026 alone, roughly 44% of everything it has ever raised.

Now line that up against the mission timeline. Independent trackers reportedly saw the satellite tumbling in May. GalaxEye publicly confirmed the loss on 7 July. The ₹49 crore cheque cleared on 10 June — in the window in between.

We don't know what the investors knew or when. Preference share allotments are usually the final legal step of a round negotiated months earlier, so this may simply be paperwork catching up to a decision taken pre-launch. But whichever way you read it, the balance sheet fact is the same: GalaxEye walked into its worst quarter with the fullest bank account in its history.

Which is exactly why it could go shopping in August instead of going into survival mode.

Which brings us to last week.

On 10 August 2026, GalaxEye announced it had acquired StarOps, a Bengaluru-based spacecraft engineering company.

On the surface this looks like a random tuck-in. It isn't. Look at what StarOps actually brings.

StarOps was incorporated in 2022, and it traces its lineage to TeamIndus — the Indian outfit that spent years chasing the Google Lunar XPRIZE with a Moon lander and rover. Several StarOps engineers worked on that programme, and the team is backed by a panel of ISRO veterans. Their expertise spans propulsion, avionics, flight computing, guidance and navigation, structures, electrical systems, mission operations and ground ops.

Crucially, they've built qualified satellite bus platforms in the 50 kg, 150 kg and 250 kg classes — with over 66% indigenisation.

That word "bus" matters. In a satellite, the payload is the interesting bit — the camera, the radar, the thing that does the job. The bus is everything else: power, thermal control, attitude control, propulsion, the flight computer, the radio. The boring plumbing. GalaxEye's genius was always in the payload. The bus is where a solar storm takes out a critical subsystem and ends your mission.

So a company that just lost a satellite to a subsystem failure has gone out and bought a team that specialises in exactly those subsystems — with a proven, largely home-grown platform to build on.

GalaxEye had already signalled this direction in July, saying it would pull more of its supply chain, component sourcing and system integration in-house. Buying StarOps is that statement turned into an org chart. Instead of assembling a satellite from vendors around the world and hoping the integration holds, GalaxEye now wants to design, build, qualify and operate the whole thing itself.

The plan from here: two new OptoSAR satellites within 24 months.

This is the part that matters if you've seen GalaxEye shares floating around on unlisted platforms and wondered whether to click buy.

Start with the number. GalaxEye's last primary round was done at roughly ₹489 crore. In the unlisted market, the stock has been quoted at something in the region of 3x that level.

Sit with that for a second. You would be paying three times what Infosys, Rainmatter, Speciale Invest and a former Defence Secretary's fund paid — investors who sit on the board, see the monthly numbers, and had access to the mission data. And you'd be paying that premium after the company's only satellite stopped responding. But you should be very clear about what you are actually buying: a queue ticket. And here's how long the queue is.

Step 1 — Wait for the next satellite. Company guidance is two new OptoSAR satellites within 24 months. Space timelines slip. Add the StarOps integration, a redesign informed by the Drishti post-mortem, and a launch slot, and 2028 is a fair working assumption.

Step 2 — Wait for it to survive. Launch isn't the milestone. Commissioning is. The spacecraft has to power up, stabilise, deploy, pass its early-orbit phase, and then keep working — the exact stretch where Drishti died. Only then does OptoSAR get proven for the first time.

Step 3 — Wait for data. Earth observation isn't a product on day one, it's an archive. Customers want historical baselines, consistent revisit rates and validated accuracy before they build workflows on top of you. That's months of imaging after the satellite is healthy.

Step 4 — Wait for order flow. And this is the slowest gate of all. GalaxEye's biggest buyers are meant to be defence, ISRO's ecosystem and government agencies. Government procurement in India runs on evaluation cycles, empanelment, pilot contracts and annual budget allocations. Even a delighted customer doesn't turn into a purchase order quickly. The NSIL reselling agreement helps distribution, but it doesn't compress a procurement calendar.

Stack those up and you're looking at a realistic four-to-six year horizon before you can even judge whether the thesis worked. And you're paying a 3x premium today for the privilege of waiting.

None of this makes GalaxEye a bad company. The technology is genuinely differentiated, the team has now bought exactly the capability it lacked, and India's sovereign demand for all-weather imagery is real and growing. The question isn't whether GalaxEye is interesting. It obviously is.

The question is whether the price in the unlisted market reflects a company that has proven its technology in orbit, or one that hasn't yet.

Right now, it's the second one.

This is an explainer, not investment advice. Unlisted shares are illiquid, thinly traded and can be marked down sharply. Do your own diligence, verify current quotes independently, and speak to a registered advisor.

Three things.

1. One, deep tech has a brutal proving ground. A SaaS startup that ships a bug pushes a patch on Tuesday. A space startup that ships a bug watches years of work tumble silently overhead, permanently out of reach. GalaxEye had raised ₹212 crore in total, and a meaningful slice of it went into a spacecraft that stopped answering the phone after about six weeks. There's no iteration cycle up there. You get one shot per satellite, and you find out whether you were right only after it's too late to change anything.

2. Two, vertical integration is a hedge, not a guarantee. Owning your supply chain gives you control over design, testing and timelines — real advantages when reliability is the product. But it also means more capital, more headcount and more things to be responsible for. And it should be said plainly: an acquisition doesn't prove a satellite will survive its next solar storm. It improves the odds. That's all any of this ever does.

3. Three, the failure isn't the end of the story — the response is. India now has 200-plus private space companies, and funding in the sector jumped 94% to about $157 million in 2025. That ecosystem is going to lose more satellites, not fewer, as it attempts harder missions. What separates the companies that survive is whether a failure produces a post-mortem and a plan, or a press release and a pivot.

GalaxEye lost its first satellite. Then it went and bought the capability it decided it was missing. In an industry where the Sun can end your company in an afternoon, that's about as good a Plan B as exists.

Until next time.

1. What is GalaxEye? GalaxEye is a Bengaluru-based space technology startup founded in 2021 by five IIT Madras alumni. It builds Earth observation satellites using its proprietary OptoSAR technology, which combines synthetic aperture radar and optical imaging on a single spacecraft.

2. What is OptoSAR technology? OptoSAR is GalaxEye's multi-sensor imaging system that puts a SAR radar sensor and a multispectral optical sensor on the same satellite. Radar sees through cloud and darkness; the optical camera produces readable, photo-like imagery. Fusing both in a single pass — which GalaxEye calls SyncFused — is meant to deliver all-weather, 24x7 imagery that a non-specialist can actually interpret.

3. Why did Mission Drishti fail? Mission Drishti launched on 3 May 2026 aboard a SpaceX Falcon 9. During its Launch and Early Orbit Phase, a severe geomagnetic solar storm exposed the spacecraft to elevated radiation, which GalaxEye's initial root-cause analysis says likely damaged a critical onboard subsystem. Communication became intermittent and was eventually lost. On 7 July 2026, GalaxEye said recovery was unlikely.

4. Why did GalaxEye acquire StarOps? StarOps, acquired on 10 August 2026, is a Bengaluru spacecraft engineering firm with roots in TeamIndus. It brings qualified satellite bus platforms in the 50 kg, 150 kg and 250 kg classes with over 66% indigenisation, plus expertise in propulsion, avionics, flight computing and guidance and navigation. Since Drishti was lost to a subsystem failure rather than a payload failure, the deal buys GalaxEye exactly the engineering layer it was weakest in.

5. How much funding has GalaxEye raised? Around ₹212 crore (roughly $24 million) across all allotments since 2021, with about ₹93 crore of that raised during 2026 alone. Its most recent primary round was priced at roughly ₹489 crore.

6. Is GalaxEye planning an IPO? No IPO has been announced. GalaxEye remains a private company, and its shares trade only in the unlisted market.

7. Should you buy GalaxEye unlisted shares? That's a personal decision, but be clear on the timeline: a new satellite launch, successful commissioning, a period of data collection, and then government and defence order flow — realistically a multi-year wait, against a current unlisted quote said to be well above the company's last funding valuation. This article is an explainer, not investment advice.