In 2015, two engineers walked out of India’s space programme convinced that the country’s most advanced material was being made in one of its most primitive ways: by hand. Carbon fibre goes into rockets, fighter jets and drones, yet the layup was still done by people with rollers and moulds. A decade later, the company they built, Fabheads Automation, says it is one of only seven firms in the world running commercial carbon-fibre 3D printing, and in June 2025 it closed a $10 million Series A led by Accel to prove that claim at scale.
Fabheads is not a consumer brand and it does not chase headlines. It is a deep-tech hardware company, incubated at IIT Madras, that automates a process most Indian factories still treat as a craft. This deep dive traces how a National Startup Award winner with barely ₹10 crore of disclosed revenue convinced Accel, Trifecta Capital and veteran investor Vijay Kedia to back a bet on machines that make the machines of the future.
Quick facts
| Company | Fabheads Automation Private Limited (CIN U74999TN2015PTC103348) |
| Founded | Incorporated 18 December 2015, Chennai, Tamil Nadu (as per MCA/Tofler) |
| Founder(s) | Dhinesh Kanagaraj and Abhijeet Rathore, both former ISRO engineers |
| Businesses | Carbon-fibre 3D printers (FabMachine), continuous fibre / Adaptive Tow Placement technology, plus end-to-end composite design, production and certification services |
| Latest FY revenue | Under ₹10 crore for FY25 (year to 31 March 2025), as classified by Tracxn; exact audited figure not public |
| Latest FY profit/loss | Not publicly disclosed; audited P&L sits behind MCA paywalls (loss-making, per the funding narrative) |
| Listed | Private (venture-backed) |
| Market value / last valuation | Valuation not disclosed; total raised reported at about $13 million (CB Insights lists $14.5 million) |
| Key shareholders / CEO | CEO Dhinesh Kanagaraj; investors include Accel, Trifecta Capital, Inflection Point Ventures, Keiretsu Chennai, Rockstud Capital, O2 Angels and Vijay Kedia |
What they do
Fabheads builds the machines and software that automate carbon-fibre and composite manufacturing, and it also runs a services arm that designs, prototypes, produces and certifies composite parts for other companies. The pitch, in one line: take a material that is expensive and slow because it is made by hand, and make it cheaper and faster by making it by machine.
- Core hardware: the FabMachine range of carbon-fibre 3D printers, described by the company and multiple trade outlets as Asia’s first and only indigenously developed carbon-fibre 3D printers.
- Core process: a Continuous Automated Fiber (CAF) 3D printing platform and an Adaptive Tow Placement (ATP) method that lay down continuous carbon fibre automatically, rather than the manual layup used in most Indian composite shops.
- Services: design, prototyping, production and certification of finished carbon-fibre components, so customers can buy parts as well as machines.
- Company-stated performance: up to 50% lower manufacturing cost, about 20% less material waste, up to 30% faster production cycles and reduced rejection rates (company figures, cited by Inc42, YourStory and Accel).
- Sectors served: aerospace, defence, mobility and electric vehicles, clean energy, shipping and drones.
The origin
The founding insight came from inside a rocket programme. Dhinesh Kanagaraj, an aerospace engineer, and Abhijeet Rathore, a mechanical engineer, met as engineers at the Indian Space Research Organisation, where Kanagaraj worked on the GSLV MkIII launch vehicle and its flow-control components. Their work depended on carbon-fibre parts, and they saw up close how those parts were made: skilled technicians, by hand, part by part, with the waste and inconsistency that manual work always carries.
In Kanagaraj’s telling, the problem was not that the engineers were bad but that the method was old. “Having experienced the limitations of manual composite manufacturing firsthand during our time at ISRO, we knew an innovative disruption was necessary, not incremental,” he has said. The pair left in 2015, incorporated Fabheads that December, and took the idea into the IIT Madras Incubation Cell, betting that a country building its own rockets and fighter jets would eventually need to make composites at industrial speed rather than artisan pace.
The struggle years
Deep-tech hardware is slow, and Fabheads had no shortcut. Building a carbon-fibre 3D printer from scratch meant years of R&D before there was a product to sell, and the early record is one of grants and small cheques rather than revenue.
- Grant-stage survival: the company took ₹10 lakh from the IIT Madras Incubation Cell and won ₹6 lakh from the DRDO Dare to Dream awards, the kind of small, non-dilutive money that keeps a hardware team alive before product-market fit (per GoodTechGo).
- Building the machine, not the market: because the core bet was on developing continuous-fibre 3D printing indigenously, most of the early years went into engineering the machine rather than booking sales, a classic deep-tech cash trap.
- Thin financials for years: even by FY23 (year to 31 March 2023), operating revenue was classified only in the ₹1 crore to ₹10 crore band, and Tofler recorded a roughly 32% fall in net worth that year, consistent with a company still spending ahead of income.
- Small paid-up capital: with paid-up capital of about ₹1.14 crore against authorised capital of ₹4.05 crore (per Tofler), Fabheads was run lean, not lavishly funded, through its build-out.
The turning point
The turn came in two steps, five years apart, and both were external validations that the machine actually worked. The first was recognition. In 2020 Fabheads won a National Startup Award in the Industry 4.0 category under the 3D printing theme, chosen from roughly 180 contestants in the category and 12 in the 3D printing field. For a company with almost no revenue, a national award was proof that the technology, not just the story, stood out.
The second, larger turn was capital and customers arriving together. In June 2025 Fabheads raised a $10 million Series A led by Accel, its biggest cheque by an order of magnitude, and did so while naming industrial customers such as the Tata Group, the Motherson Group and ISRO, alongside a defence development project for the Indian Air Force and installations for space startup Agnikul Cosmos and electric-aircraft firm ePlane. The contrast is the story: from ₹8 crore pre-Series A money in 2021 to a $10 million round in 2025, roughly a ten-fold jump in a single round, funded on the strength of machines now running in real factories.
The money behind it
Fabheads has raised about $13 million in total across pre-seed, angel, pre-Series A and Series A rounds (per Inc42, YourStory and Indian Startup News); CB Insights lists a higher $14.5 million. The shape is typical deep-tech: years of small, patient money, then one institutional round once the hardware was proven.
- Pre-seed, about $1 million: the founders started with roughly $1 million in pre-seed funding (per Accel), enough to build early prototypes.
- Pre-Series A, ₹8 crore (May 2021): led by Inflection Point Ventures, with existing backers Keiretsu (Chennai chapter) and investor Vijay Kedia of Kedia Securities participating (per YourStory).
- Series A, $10 million (June 2025): led by Accel, with Trifecta Capital providing about $2.3 million as venture debt within the round; reported in rupee terms at roughly ₹83.5 crore to ₹86.7 crore at the time (per Indian Startup News and Inc42).
- Other named investors: Rockstud Capital and O2 Angels also feature on the cap table (per Inc42).
What each round changed: the grants and pre-seed paid for the R&D; Inflection Point’s ₹8 crore in 2021 let Fabheads move from prototype to product and add customers; Accel’s 2025 round is explicitly a scale-up cheque, earmarked for an 80,000 to 100,000 sq ft factory in Bengaluru’s Aerospace Park, more than 70 additional manufacturing machines, and a bigger engineering and leadership team, targeted to be operational within about six months of the raise.
How it makes money
Fabheads has two ways to earn, and they sit at different points on the risk curve.
- Selling machines (capital sales): the FabMachine 3D printers are sold to manufacturers and institutions. IIT Madras, for example, stood up an additive-manufacturing lab with 25-plus FabMachines, a single deployment that shows what a machine order looks like. Machine sales are lumpy but high-value.
- Selling parts and services (recurring-ish revenue): the services arm designs, prototypes, produces and certifies composite components for customers who want the parts, not the printer. This is where names like Tata, Motherson, ISRO, Agnikul and ePlane come in as buyers of output.
- Where the margin sits: the company’s own case for margin is the cost stack it removes: up to 50% lower manufacturing cost, about 20% less material waste and up to 30% faster cycles versus manual layup (company-stated). If those hold at volume, the machine pays for itself in scrapped material and labour saved.
- The moat it claims: more than 10 patents across seven Asian and European countries, and a position as one of only seven companies globally with commercial carbon-fibre 3D printing operations (company-stated, via Inc42 and Accel).
The part people get wrong: Fabheads is not really a “3D printing” company in the desktop-printer sense. It is an industrial automation company whose product happens to be a fibre-placement robot. The money is meant to come from displacing manual composite lines, not from hobbyist printing.
The numbers
Fabheads is a private company that does not publish audited accounts openly, and the granular MCA filings sit behind paid databases. What is public are revenue bands and one net-worth movement, so the honest table below shows the disclosed ranges rather than invented precision. Figures are as classified by Tracxn and Tofler; unit is ₹ crore.
| Financial year | Operating revenue (₹ crore) | Profit / loss |
| FY23 (to 31 Mar 2023) | ₹1-10 crore band (Tofler); net worth down about 32% YoY | Not publicly disclosed |
| FY24 (to 31 Mar 2024) | Not publicly disclosed | Not publicly disclosed |
| FY25 (to 31 Mar 2025) | Under ₹10 crore (Tracxn classification) | Not publicly disclosed |
The clearer numbers are on the capital side, and they tell the same story a revenue table would: a company still small on the top line but heavily backed on the balance sheet.
| Round | Date | Amount | Lead / key backers |
| Pre-seed | Early stage | About $1 million | Angel / early backers |
| Pre-Series A | May 2021 | ₹8 crore | Inflection Point Ventures; Keiretsu Chennai; Vijay Kedia |
| Series A | June 2025 | $10 million (about ₹83.5-86.7 crore) | Accel (lead); Trifecta Capital (about $2.3 million venture debt) |
| Total raised | To date | About $13 million (CB Insights: $14.5 million) | Multiple |
Where the money comes from
Fabheads does not publish a segment or geography split, so this section describes the demand mix its own disclosures and customer names point to, attributed as such.
- By sector: aerospace and defence sit at the core, with a defence development project for the Indian Air Force and space customers ISRO and Agnikul Cosmos named publicly; mobility and EVs (ePlane, electric aircraft), clean energy, shipping and drones are the growth adjacencies (per Accel, Inc42 and Indian Startup News).
- By customer type: a mix of large industrials (Tata Group, Motherson Group) buying parts and capability, and deep-tech startups (Agnikul, ePlane) buying components for their own hardware.
- By product line: capital equipment (FabMachine printers, such as the 25-plus units at IIT Madras) on one side, and design-to-certification services on the other.
- The surprise: for a “3D printing” company, the pull is coming from heavy, safety-critical industries such as aerospace and defence rather than consumer or prototyping markets, which is why certification services matter as much as the printer itself.
The risks
- Revenue is still tiny against the ambition: with revenue classified under ₹10 crore in FY25 while raising a $10 million round, Fabheads is being funded on promise, not cash flow. If the Bengaluru factory and 70-plus machines do not convert into orders quickly, the gap between spend and sales widens, and the FY23 net-worth erosion (down about 32%, per Tofler) shows how fast a hardware balance sheet can thin.
- Customer concentration in slow, cyclical buyers: aerospace, defence and space are long-cycle, tender-driven and lumpy. A single delayed defence programme or a paused space customer can move a quarter for a company this size, and the marquee names (Tata, Motherson, ISRO) are exactly the kind that buy slowly and negotiate hard.
- Deep-tech execution and competition: being “one of seven globally” cuts both ways; the other six are well-capitalised international players, and scaling precision fibre-placement machines from a lab into a 100,000 sq ft factory is an execution risk in itself. Patents (10-plus across seven countries) help, but hardware moats erode if rivals ship faster.
The takeaway
The transferable lesson from Fabheads is about patience with the right kind of problem. The founders did not pick a market that would reward them quickly; they picked one where the incumbent method was clearly wrong and clearly hard to fix, then spent years and grants building the fix before a single large customer showed up. Deep-tech does not let you fake traction, so the discipline is to survive on small, non-dilutive money long enough for the technology to become undeniable, and to let a national award and named industrial customers do the persuading that a pitch deck cannot. The reward, when it comes, is a round that jumps ten-fold in one step, because by then the risk that remains is execution, not invention.
Frequently asked questions
What does Fabheads Automation do?
Fabheads builds carbon-fibre 3D printers and continuous-fibre placement machines that automate composite manufacturing, and it also offers design, production and certification services for carbon-fibre parts. Its customers span aerospace, defence, mobility, clean energy and drones.
Who founded Fabheads and when?
It was founded by former ISRO engineers Dhinesh Kanagaraj and Abhijeet Rathore, and incorporated on 18 December 2015 in Chennai. It was incubated at the IIT Madras Incubation Cell.
How much funding has Fabheads raised?
About $13 million in total (CB Insights lists $14.5 million), including a ₹8 crore pre-Series A in May 2021 led by Inflection Point Ventures and a $10 million Series A in June 2025 led by Accel, with Trifecta Capital participating via venture debt.
What is Fabheads best known for?
It is described as Asia’s first and only maker of indigenously developed carbon-fibre 3D printers, and it won a National Startup Award in 2020 in the Industry 4.0 / 3D printing category. It says it is one of only seven companies globally with commercial carbon-fibre 3D printing operations.
Is Fabheads profitable?
Fabheads does not publish its accounts openly, and detailed profit or loss figures are not public. Its disclosed revenue was under ₹10 crore in FY25, and it is being funded to scale rather than on current profits, so it should be assumed to be investing ahead of income.
Sources
Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics); note that the 2025 Series A is cited at its reported rupee value at the time, not re-converted.
- Inc42, “Fabheads Bags $10 Mn To Manufacture Composite Materials” (June 2025)
- YourStory, “Fabheads raises $10M in Series A led by Accel” (June 2025) and “Carbon-fiber startup Fabheads raises Rs 8 crore” (May 2021)
- Accel, “Our Investment in Fabheads” note (2025)
- Indian Startup News, “Chennai-based deeptech startup Fabheads raises $10 million” (June 2025)
- Tofler, Fabheads Automation Private Limited company page (CIN U74999TN2015PTC103348), accessed September 2026
- Tracxn, Fabheads Automation legal-entity and company profile, accessed September 2026
- GoodTechGo, “Fabheads Automation wins national startup award for 3D printing” (2020 award)
- CB Insights and Crunchbase, Fabheads Automation funding profiles, accessed September 2026
Found an error? Write to us and we’ll correct it in the open, dated, on the piece.

