India’s Deep-Tech Startups to Grow Fivefold — Vish Sahasranamam (ThePrint)

ELI5/TLDR

Vish Sahasranamam runs Forge, an incubator network in Tamil Nadu that helps hardware-plus-software startups — what he calls “phygital” — actually get their products into factories, warships, and supply chains. His pitch: India’s 25,000 deep-tech startups are going to 5x in a decade because the government finally started buying from them (especially defense, via a program called iDEX), and because global companies fleeing China want to manufacture in India anyway. The central idea he keeps circling is that every company — cement, banks, defense, PSUs — has to stop thinking of itself as whatever-it-makes and start thinking of itself as a tech company in that business. Otherwise the small, fast startups eat its lunch.

The Full Story

Who is Forge and what does “phygital” mean

Forge Innovation and Ventures, based in Coimbatore with seven other centres in Tamil Nadu, calls itself India’s largest “open innovation network” for phygital startups. Phygital here means hardware-first, digital-plus — the sensor, the motor, the drone, the medical device, not another SaaS dashboard. Forge runs labs where founders go from prototype to production-ready unit, and it brokers the awkward introductions between industrial buyers (factories, defense PSUs, port authorities) and the scrappy startup that claims it can do the thing cheaper.

Their three pillars are talent, technology, and ventures. The unspoken thesis is that a SaaS startup goes innovation → commercialisation, but a hardware startup has to cross an additional moat: innovation → production → industrialisation → commercialisation. Industrialisation is the step where your axial flux motor has to pass the OEM’s type approvals and design acceptance gates before a single unit gets ordered. This is where most hardware startups quietly die, and it’s what Forge exists to un-stick.

The bigger story: India graduating from user of tech to builder of tech

Sahasranamam’s macro frame is that India’s first 50 years of tech — IT services, fintech, e-commerce — were about applying imported technology to local problems. Flipkart, PhonePe, Ola. All built on top of mobile infrastructure someone else invented. He points out, correctly, that India didn’t achieve population-scale digital banking because the RBI ordered it. Startups did it.

The last seven or eight years, he argues, something different is happening: Indian researchers are crossing the line from paper-and-patent into actual companies. Bio-pharma clusters in Hyderabad are spitting out startups instead of just journal articles. Aerospace, defense, life sciences — same template. A scientist teams up with a college friend who went into industry and learned marketing, venture money shows up, and a research project becomes a company.

“I don’t want to just put a paper out. I don’t want to put just an IP patent out. I actually want to see if I can translate that into a commercial product.”

His number: roughly 25,000-30,000 deep-tech startups in India today, and he thinks there’s room for 5-10x in a decade.

iDEX and the defense buying revolution

The most concrete claim in the interview is about iDEX — the Innovations for Defence Excellence program — which Sahasranamam helped architect from 2018-19. The point of iDEX was to convince the military and defense PSUs (HAL, BEL, and the 16 companies carved out of the Ordnance Factory Board) to actually buy from startups rather than only from the old PSU behemoths.

The numbers he cites: around 600 iDEX projects running, roughly ₹2,500-3,000 crore of orders approved for startups that have cleared trials, and ₹1,200-1,500 crore already actually issued as orders. A startup-to-defense pipeline that ships. He mentions new programs — the 14th Defense India Startup Challenge, Aditi 4.0, and a new one called Drishti for DPSU-specific innovation challenges. A sister program, S2i2, tries to do the same for the maritime sector under the Ministry of Ports, Shipping and Waterways.

Why Indian MSMEs are about to get an ultimatum

The interesting thread here is about the thousands of small manufacturers in Hosur and similar industrial towns. They supply the big OEMs (TVS, Ashok Leyland, Titan, Tata Electronics), and they’re about to be squeezed from three directions at once: rising quality demands, rising wage costs, and — this is the one people don’t talk about enough — net-zero compliance.

He quotes a former Volvo India head saying that within five to seven years, MSME suppliers in India have to hit minimum sustainability targets set in Sweden. If they can’t, Volvo moves the supply chain to Thailand or Indonesia. The MSMEs have two problems: they are organised for production, not innovation, and they cannot afford to import German automation equipment. Sahasranamam’s line on this is the memorable one:

“How long are we going to import the 800-lb gorilla from Germany or Japan? We need that 15-lb chimpanzee, but you cannot get it out of chipping and chopping the 800-lb gorilla. You’ve got to build that 15-lb chimpanzee from the ground up.”

Translation: frugal innovation isn’t just stripping features off expensive equipment. You have to build the cheap version fresh, from local first principles.

Tesla, Amazon, and the cement company that has to become a tech company

The strongest intellectual move of the interview is a fairly standard one, done well. Sahasranamam traces how technology’s role in business has mutated through four stages: enabling the business (old IT), operating the business (Amazon ordering), reshaping the business (AWS as 40% of Amazon’s market cap), and finally becoming the business (Tesla, whose peak valuation was greater than the top 20 automakers combined).

He invokes the Jamie Dimon line: “We’re a technology company with a banking license.” His point is that Indian PSUs, and specifically the old Navaratnas and Mini Ratnas — railways, defense, banks — have to internalise the same mental shift or they’ll get outflanked. And then he applies it at the firm level: UltraTech will not win because it has the best gypsum mines. It’ll win because it has AI tech in some corner of its value chain that nobody else can match.

This isn’t novel, but it’s a clean version of the argument, and it’s the right one to keep repeating to people who still think of “tech” as the IT department.

Returning engineers, China-plus-one, and the trigger theory

When asked whether Indian-origin engineers in the US are likely to come home, his answer is cautiously optimistic. He thinks the migration has already started — dozens, not thousands — and that the coming wave won’t be mass repatriation but trigger-driven: one delayed promotion, one anti-H1B executive order, one personal crisis, and someone’s on the next flight to Coimbatore. He tells a story about a Chennai-origin founder running a climate tech startup in Perth who moved lock-stock-and-barrel to Coimbatore after Sahasranamam walked him through the cost-of-innovation math (roughly 1/5 in India versus Australia).

On China: he argues the comparison is almost impossible to run cleanly because China’s innovation ecosystem is opaque, whereas India’s is embarrassingly transparent — walk into any lab, read any filing. He thinks that transparency is actually an asset for inbound multinationals doing due diligence.

The regulator question

The closing chapter is the most practically interesting. The host asks: what do you do about Indian regulators in an AI-sweeping-everything world? Sahasranamam drops a meme statistic — telecom companies capture only 0.03% of the ChatGPT economy they carry the traffic for — and argues that telecom is the cautionary tale. Airtel didn’t engage with startups when it was printing money. Now it’s stuck being dumb pipes while OpenAI and Perplexity take the value.

He wants regulators to evolve through five Cs: clamping down, compliance, cooperation, co-creation, and finally championing — where a regulator actively writes playbooks, runs sandboxes, and nudges the sector forward. He thinks the ones to watch are mining, oil and gas, and power, which have the heaviest regulatory overhang and the least engagement with startups.

“The days of clamping things down is way behind us. But if we can just get them to be a little more aspirational, maybe visionary, I think that’s the best that can happen for India.”

The interview ends on expansion plans: Pune and Vizag are next. Vizag especially — he wants an innovation hub inside the port itself, because the port is sitting in the middle of an industrial cluster of fertiliser, petrochemicals, refineries, and now ArcelorMittal steel.

Claude’s Take

This is a solid, substantive interview, but it’s basically a founder doing a victory lap, and you should read it as such. Sahasranamam is a real operator — iDEX is genuinely one of the better things to happen to Indian defense procurement in years, and Forge seems to be doing real work — but he has every incentive to paint the deep-tech boom as bigger and more inevitable than it is.

The strongest parts are the concrete stuff. The iDEX numbers check out with what’s publicly reported, give or take. The Volvo net-zero story is plausible and directionally correct — European OEMs are genuinely doing this to their supply chains. The “phygital valley of death between prototype and industrialisation” observation is real and underdiscussed in Indian startup coverage, which skews heavily toward consumer apps. The “15-lb chimpanzee, not a chopped-up 800-lb gorilla” line is a better version of “frugal innovation” than most people articulate.

The weaker parts are the extrapolations. “25,000 deep-tech startups growing 5x in a decade” — depends entirely on how loose your definition of deep tech is, and in India’s ecosystem people count anything with sensors. The claim that Indian-origin US engineers will return in meaningful numbers is a perennial optimistic forecast that has been wrong for fifteen years; maybe this time is different, but the evidence he offers — “a few hundred over the next two years” — is not actually a trend, it’s a rounding error on a diaspora of millions. The China comparison gets handwaved: saying China is “opaque” so we can’t really compare is technically true but also conveniently lets him skip over the fact that China’s deep-tech ecosystem genuinely is further along, especially in electronics, EVs, and advanced manufacturing.

The “every company is a tech company” framing is useful but also not new — it’s been standard McKinsey-deck material for a decade. What’s actually harder and more interesting is the question nobody asks him: what happens when the startup itself becomes the OEM and starts wanting to protect its own margins? That’s the stage Tesla is in. India doesn’t really have that yet, and the path from “ecosystem of plucky phygital startups” to “globally competitive vertically integrated hardware companies” has a lot of dead bodies along it.

The AI jobs-displacement section is the weakest. His analogy — bank tellers displaced by ATMs didn’t cause mass unemployment — is a commonly cited and commonly wrong use of the Bessen finding. ATMs were labour-augmenting; it’s not clear AI is. And his follow-up — “maybe a displaced factory worker becomes a Swiggy delivery rider and it might even be more fun” — is cheerful in a way that suggests he has not thought about this very hard.

The regulator section is the one with teeth. The 0.03% telecom number is a good stat, probably roughly right in direction even if the exact figure is hand-wavy, and the argument that Indian regulators in mining, power, and oil-and-gas have completely abdicated their role as innovation champions is correct and specific. This is the stuff I’d want to hear an hour more of.

claude_score: 6. Decent. It’s an informative founder interview with several genuinely useful data points — iDEX numbers, the MSME net-zero squeeze, the telecom cautionary tale — wrapped in a lot of standard bullishness. The host asks reasonable questions but doesn’t push on any of the soft spots. Worth reading if you care about Indian industrial policy or hardware startups; skippable if you’re looking for frontier ideas. Solidly middle-of-the-road.

India's deep-tech startups to grow fivefold from the current 25,000