Startup Deep Dive : TakeMe2Space — it raised $5 Mn then lost its satellite four days later

On 8 January 2026, a Hyderabad startup called TakeMe2Space closed a $5 million round to build a constellation of satellites that think for themselves. Four days later, on 12 January 2026, the rocket carrying its first flagship satellite broke apart mid-flight, and the satellite it had just raised money to multiply was gone before it ever switched on.

That is the whole tension of TakeMe2Space in two dates: a founder betting that satellites can be rented out like cloud servers, and the oldest problem in the space business — getting to orbit at all — proving him right about the risk before it proved him right about the business. What survives that contradiction, and how, is the rest of this story.

Quick facts

What they do

TakeMe2Space builds small satellites that carry AI processors instead of just cameras and radios, and it rents out the compute on those satellites the way a cloud provider rents out a server. Its flagship platform, OrbitLab, lets researchers, universities, defence users and enterprises upload their own AI models and run them directly in orbit, on data the satellite has just captured, instead of beaming raw imagery down to Earth and processing it in a data centre. The company also sells the hardware it had to invent to make that possible — radiation-hardened compute modules under the RadShield name, and a compact satellite battery pack called PowerBank-50 — to other satellite builders who don’t want to build their own from scratch. It is a Hyderabad company, headquartered out of the Telangana space-tech ecosystem, selling primarily to Indian and international research institutions, universities and early commercial customers.

The origin

Ronak Kumar Samantray is not new to watching a good idea get strangled by infrastructure cost. He had already co-founded NowFloats, a SaaS platform for small and medium businesses, in 2012, alongside Jasminder Singh Gulati, Nitin Jain and Neeraj Sabharwal; Reliance Industries acquired it on 12 December 2019, folding it into the Jio ecosystem, as reported by YourStory at the time. It was while running NowFloats through the pandemic, according to Inc42’s account of the founding story, that Samantray watched the company’s revenue fall by roughly 70% while its AWS hosting bill fell by only about 10% — a mismatch between what the business earned and what the infrastructure still cost, whether it was used or not. That gap between demand and fixed infrastructure spend stayed with him.

The specific idea for TakeMe2Space came from a more literal frustration: Samantray wanted to launch his own code into space, and found there was no accessible way in India for an individual or small team to rent time on an orbiting satellite the way one could rent a cloud server on Earth, according to ThePrint’s ground report on the company. Unable to buy that access, he decided to build the satellite instead. He took the idea to T-Works, the Telangana government’s hardware prototyping lab in Hyderabad, where Anand Rajagopalan, Sharat Reddy and Firoz Ahammed — later TakeMe2Space’s core team — were working before they joined the venture full time.

The struggle years

The company’s early years were less a single dramatic failure than a string of postponed dates and a team learning space hardware with no institutional safety net. Three episodes stand out, unsoftened, in the public record.

First, there was no existing satellite in India that an outside team could simply buy time on, which is what forced TakeMe2Space to become a satellite manufacturer rather than only a software company — a far more capital- and time-intensive path than the founders originally set out on, per ThePrint’s reporting. Second, thermal management in orbit turned out to be, in the company’s own words to ThePrint, “non-trivial”: keeping AI processors cool without the convective air that cools a server on Earth is a genuinely different engineering problem, and it pushed back the timeline for a fully operational satellite. ThePrint’s ground report, filed while the company was targeting a launch on ISRO’s SSLV “by September 2025”, shows that date itself was already a slip from earlier ambitions; the company’s actual maiden flagship satellite, MOI-1, did not fly until the PSLV-C62 mission in January 2026, months past that window. Third, and most costly, was the launch failure itself, covered in the next section — TakeMe2Space had, by its own admission to ThePrint after the fact, left most of its payloads uninsured, a decision common among early-stage space startups because of the cost of coverage, but one that turned a rocket anomaly into a full write-off of hardware the team had spent months building.

• Had to become a hardware and satellite-manufacturing company to solve what began as a software access problem (ThePrint ground report)
• Missed its own stated target of a first satellite launch “by September 2025” on ISRO’s SSLV (ThePrint ground report)
• Flew its flagship MOI-1 satellite without insurance on most of the payload, a gap exposed when the January 2026 launch failed (ThePrint, May 2026)

The turning point

The clearest before-and-after moment in TakeMe2Space’s short history sits across four days in January 2026. On 8 January 2026, Inc42 reported that the company had closed a $5 million seed round led by Chiratae Ventures, with Artha Venture Fund, Seafund and Unicorn India Ventures also participating, explicitly to expand its satellite constellation to six spacecraft and add roughly 5 kW of in-orbit compute capacity. On 12 January 2026, ISRO’s PSLV-C62 mission suffered a third-stage anomaly shortly after liftoff from Sriharikota, and lost thrust before it could place its payloads in orbit. The mission was carrying the EOS-09 earth observation satellite as its primary payload and 15 co-passenger satellites, including TakeMe2Space’s own MOI-1 — its maiden flagship satellite and the very hardware the fresh funding was meant to multiply — along with payloads from Dhruva Space, EON Space Labs, OrbitAid and student teams from four countries, as reported by ThePrint.

The numbers on each side of that week are stark: a company had just raised outside capital on the premise that it could scale a working satellite business, and within days its only flagship satellite in the pipeline was destroyed, with little to no insurance to cushion the loss. What happened next is arguably more telling than the failure itself. Because TakeMe2Space had built backup subsystems during MOI-1’s original assembly, it was able to begin rebuilding a near-identical satellite, MOI-1a, almost immediately. By April 2026, Analytics India Magazine reported the company had booked a slot on SpaceX’s Falcon 9 Transporter-18 rideshare mission; by May 2026, ThePrint reported that launch was targeted for no earlier than October 2026. Anand Rajagopalan, the company’s business strategy lead, told ThePrint plainly: “Any space startup is burning cash whether or not they launch. Accessing space-based services quickly matters for business” — an admission that the meter runs on payroll and overheads regardless of whether a rocket ever leaves the pad.

The money behind it

TakeMe2Space has raised money in two disclosed institutional rounds, plus a separate, milestone-linked government grant programme, and has not disclosed a valuation at any stage.

• Pre-seed, March 2025: ₹5.5 crore, led by Seafund, with Blume Ventures, Artha Venture Fund and AC Ventures participating (Seafund’s own release and multiple Indian startup-news outlets, 4 March 2025); Inc42 DataLabs separately records this round at $629,500
• Seed, 8 January 2026: $5 million, led by Chiratae Ventures, with Artha Venture Fund, Seafund and Unicorn India Ventures participating (Inc42, 8 January 2026), earmarked for expanding the satellite constellation to six spacecraft and scaling in-orbit compute
• Government grant, disclosed June 2026: TakeMe2Space was named one of the first three spacetech startups funded under IN-SPACe’s Technology Adoption Fund (TAF) scheme, alongside Astrobase Space Technologies and SatSure Analytics; the scheme caps investment at ₹25 crore (about $2.6 million at $1≈₹96.0, 18 September 2026) per project, with milestone-linked disbursement, though TakeMe2Space’s own disbursed amount was not disclosed in Inc42’s report
• Total raised: $5.63 million across the two institutional rounds, as of July 2026 (Inc42 DataLabs)

What each backer is reported to have changed: Seafund led the earliest institutional check when the company had little more than a prototype, which Inc42’s coverage frames as validation for a hardware-heavy, pre-revenue bet; Chiratae Ventures’ larger seed round is the one tied explicitly, in the company’s own funding announcement, to the jump from a single satellite to a six-spacecraft constellation and to opening sales offices across India, the US and Australia; and the IN-SPACe grant is a government signal of technical credibility that is separate from, and additional to, private capital.

How it makes money

The pitch is usage-based compute pricing for orbit, mirroring how cloud providers charge for server time on Earth.

• Money in — compute rental: OrbitLab charges roughly $2 per minute of in-orbit AI processing time, according to Inc42’s reporting on the company; ThePrint’s earlier ground report cited a range of $2 to $4 per orbital minute, meaning a single 90-minute low-Earth-orbit pass could cost a customer in the region of $360
• Money in — hardware sales: subsystem products such as the PowerBank-50 satellite battery are sold off the shelf to other satellite builders, at a listed price of ₹1,04,900 per unit, according to trade coverage of its July 2026 flight-heritage milestone
• Costs out: satellite manufacturing and testing, launch costs (industry launch pricing runs around $1,500 per kilogram today, with the company targeting roughly $600 per kilogram as newer providers enter the market, per ThePrint), ground-station and assembly-integration-and-test (AIT) partnerships, and an expanding engineering and global sales team across India, the US and Australia (Inc42)
• Where the company says the margin sits: in processing data closer to where it is captured rather than transmitting it whole; the company claims this can cut data-transmission costs by as much as 85% and shrink turnaround time from a 24-to-72-hour ground-processing cycle to minutes, per Inc42’s July 2026 profile of the business
• The part people get wrong: that the company is competing with its own customers by building proprietary AI models. Inc42’s reporting is explicit that TakeMe2Space positions itself as infrastructure only — it does not build or sell the AI models that run on its satellites, only the orbital compute layer they run on

The numbers

TakeMe2Space is genuinely too young to show the three-to-four-year revenue and profit history this format normally carries, and this piece will not manufacture one. The company was incorporated as TM2Space Technologies Private Limited on 10 June 2024 (MCA/RoC Hyderabad filing), which means FY25 — the year ended 31 March 2025 — is its first full financial year on record. The only verifiable figure from that year is revenue; no profit or loss figure was found in any source opened this session.

Set against that thin financial base, the company’s total institutional funding of $5.63 million (Inc42 DataLabs) is more than 200 times its disclosed FY25 revenue — a ratio typical of a pre-revenue deep-tech hardware bet, where investors are pricing the constellation the company says it will build, not the compute it has sold to date.

Where the money comes from

Public reporting does not break out TakeMe2Space’s revenue by customer segment or geography in rupee or dollar terms, so this section states what is verifiable about its customer mix and pipeline rather than a numeric split.

• Customer types named in reporting: research institutions, universities, commercial enterprises and AI startups uploading models to OrbitLab (Inc42); students and academic teams specifically called out in ThePrint’s ground report
• Named pilot/commercial interest: Tiger Analytics has been reported exploring agricultural applications of the platform (Inc42, July 2026)
• Pipeline scale: ThePrint’s ground report cited a waitlist of around 40 clients for orbital compute time; Analytics India Magazine’s April 2026 report on the MOI-1a rebuild referenced roughly 15 customers lined up for that satellite’s AI-lab-in-space demonstrations
• Geographic expansion intent: the January 2026 seed round was explicitly earmarked for sales expansion into the US and Australia in addition to India (Inc42), alongside an existing partnership with Australian satellite-servicing platform HEX20 signed in July 2024 (company milestones page)
• The surprise: a meaningful share of near-term revenue looks likely to come not from renting AI compute at all, but from selling subsystem hardware — batteries and radiation shielding — to other satellite builders who need flight-proven components faster than they can build their own, evidenced by the commercial PowerBank-50 listing that followed its July 2026 flight-heritage milestone

The risks

• Launch dependency and no insurance: TakeMe2Space’s entire flagship satellite programme depends on third-party rockets it does not control. Its maiden satellite, MOI-1, was destroyed in the PSLV-C62 failure on 12 January 2026, and the company has confirmed most of its payloads carried no insurance, meaning a repeat failure would again be a full write-off rather than a claims process (ThePrint, May 2026)
• Unproven unit economics: Inc42’s own July 2026 profile of the business describes orbital computing economics as “still largely unproven” at industry level — the company is pricing compute per orbital minute before it has demonstrated that customers will pay enough, at enough volume, to cover the cost of building and launching each satellite
• Concentration on a single government-linked funding lever: the IN-SPACe Technology Adoption Fund that co-funds TakeMe2Space caps disbursement at ₹25 crore per project and ties payouts to milestones, which means delays of the kind the company has already experienced (a missed 2025 launch window, a destroyed satellite in January 2026) can directly slow the release of that capital, on top of any effect on investor confidence

The takeaway

The lesson in TakeMe2Space’s short history so far is not really about space. It is about what happens when a company’s core infrastructure bet and its core infrastructure risk are the same thing. Samantray built his first company on the insight that cloud costs don’t fall when revenue does; he built his second on the bet that orbital compute could be rented the same way. But a cloud server that fails gets replaced in minutes by a different physical machine in the same data centre. A satellite that fails is gone, along with months of assembly and, in TakeMe2Space’s case, most of the insurance that might have covered it. The company’s response — pre-built backup subsystems that let it rebuild MOI-1a and rebook a launch within roughly three months of losing MOI-1 — suggests the team has internalised that lesson operationally, even if the economics of the business it is building are still, by its own backers’ admission, unproven. For any founder building infrastructure with a single point of physical failure, the transferable point is the same: redundancy has to be built before the failure, not budgeted after it.

Frequently asked questions

What does TakeMe2Space actually sell?

It sells rented access to AI computing power carried on its own satellites, mainly through its OrbitLab platform, priced at roughly $2 to $4 per orbital minute depending on the source cited (Inc42; ThePrint). It also sells satellite hardware components, such as its PowerBank-50 battery, to other satellite builders.

Who founded TakeMe2Space and when?

Ronak Kumar Samantray, who previously co-founded NowFloats (acquired by Reliance Industries in December 2019), founded TakeMe2Space in 2023 according to the company and Inc42’s reporting; the entity was formally incorporated as TM2Space Technologies Private Limited on 10 June 2024, per MCA records.

How much funding has TakeMe2Space raised?

Two disclosed institutional rounds total $5.63 million as of July 2026 (Inc42 DataLabs): a ₹5.5 crore pre-seed round in March 2025 led by Seafund, and a $5 million seed round in January 2026 led by Chiratae Ventures. It has also been named a recipient under IN-SPACe’s government-backed Technology Adoption Fund, though its individual grant amount was not disclosed.

What happened to TakeMe2Space’s first satellite?

Its maiden flagship satellite, MOI-1, was lost when ISRO’s PSLV-C62 mission suffered a third-stage anomaly shortly after launch on 12 January 2026, destroying the primary EOS-09 payload and all 15 co-passenger satellites aboard, including MOI-1. The company is rebuilding a near-identical satellite, MOI-1a, targeted to launch on SpaceX’s Falcon 9 Transporter-18 mission no earlier than October 2026.

Is TakeMe2Space profitable?

No profit or loss figure has been disclosed in any source reviewed for this piece. Its only publicly reported financial figure is FY25 revenue of ₹21.3 lakh (Inc42 DataLabs), for the year ended 31 March 2025 — its first full financial year after incorporation in June 2024.

Sources

Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics).

• Inc42 DataLabs, TakeMe2Space company profile (funding, revenue, headcount), accessed September 2026
• Inc42, “TakeMe2Space Raises $5 Mn To Expand Its Satellite Constellation”, 8 January 2026
• Inc42, “IN-SPACe Funds Three Spacetech Startups Under Technology Adoption Fund”, 11 June 2026
• Inc42, “Can TakeMe2Space Build The AWS Of Space?”, 27 July 2026
• ThePrint, “A Hyderabad startup is launching India’s 1st ‘thinking’ satellite. A brain in space”, ground report
• ThePrint, “Hyderabad start-up TakeMe2Space turns to SpaceX after ISRO PSLV mission failure”, May 2026
• Analytics India Magazine, “TakeMe2Space Schedules MOI-1a Launch on SpaceX Falcon 9 After PSLV Setback”, 16 April 2026
• Seafund, “TakeMe2Space Raises ₹5.5 Cr for India’s Space AI Lab”, 4 March 2025
• YourStory, “Reliance acquires SaaS platform for SMEs NowFloats”, December 2019
• UNI India, “TakeMe2Space’s PowerBank-50 becomes India’s first off-the-shelf satellite battery to achieve space flight heritage”, July/August 2026
• Instafinancials (MCA-sourced), TM2Space Technologies Private Limited company overview, CIN U61900TS2024PTC186392, accessed September 2026
• Tracxn, TakeMe2Space company profile, accessed September 2026
• TakeMe2Space company website, About page (tm2.space/about-tm2s), accessed September 2026

Found an error? Write to us and we’ll correct it in the open, dated, on the piece.