India’s ₹1.5 Lakh Crore DARPA Moment For Deep Tech | Govindraj Ethiraj

ELI5 / TLDR

India has set up a single big government body to pump money into hard science and frontier technology — roughly ₹1.5 lakh crore worth. It does two different things: hands out grants to universities and labs (no strings, no payback) and makes investments in private companies (loans that wait patiently for returns). The first batch of 22 companies just got funded, ranging from 3D-printed rockets to 6G to satellites. The CEO, a former IBM and Microsoft scientist, walks through how it works and admits they’re still figuring a lot of it out as they go.

The Full Story

One body, many wallets

The organization is called the Anusandhan National Research Foundation, or ANRF. Think of it less as a charity and more as a holding company for different kinds of money. The CEO, Dr. Shivkumar Kalyanaraman, offers the cleanest analogy in the interview: when Nandan Nilekani built Aadhaar, Aadhaar was the project and UIDAI was the body that ran it. ANRF is that kind of body — a permanent, legally established institution — and inside it sit several distinct funds, each with a different job.

“ANRF is a statutory body. Within it, we have funds and you know, potentially organizations which do specific things. RDI is a fund. It’s a special purpose fund of ANRF.”

The comparison everyone reaches for is DARPA, the US defense agency famous for funding moonshots that later became the internet and GPS. The CEO accepts the comparison but sharpens it. ANRF actually plays two roles the Americans split across two agencies: the no-strings grant-giving of the National Science Foundation, and the mission-driven, “go build this specific impossible thing” style of DARPA.

Grants versus investments — the crucial split

Here is the distinction worth holding onto. ANRF spends money two ways, and they do not mix.

One way is grants — money you don’t pay back. These go to the not-for-profit world: universities, research labs, and so on. This is for early-stage, basic research where you can’t promise anyone a return.

The other way is investments — loans or convertible instruments that expect their money back eventually. These go only to private, for-profit companies. The idea is that the private sector is the engine that actually turns a lab result into something at scale.

“The private sector is the you know, vehicle to really translate scale and really create the ultimate economic opportunities. Whereas the not-for-profit sector… are at the earlier stages.”

The mental model is a pipeline: a grant seeds an idea in a lab, the idea matures, and somewhere downstream a private company picks it up with investment capital and scales it.

”Patient capital” — and what that actually means

The recent ₹1.5-lakh-crore headline came from one piece of this machine: the Research, Development and Innovation (RDI) scheme, announced by the Prime Minister in November and run day-to-day by two fund managers (TDB and BIRAC). TDB was first out the gate, funding 22 companies — names a tech-watcher would recognize: Tejas Networks (Tata group), Ather Energy, ideaForge, plus rocketry and drone startups like Agnikul, GalaxEye, and E-Plane.

The money comes as patient capital — long-term debt or convertibles, deliberately structured so founders don’t have to give away ownership. Tickets run from ₹50 crore to ₹250 crore, and RDI only ever funds half a project. So a ₹250-crore cheque means a ₹500-crore project, with the company finding the rest.

The CEO is careful about what “patient” means, because people assume it just means “lots of money for a long time.” It doesn’t. Some deep tech needs time but not much money — a small team working for years. Other deep tech, like drug discovery, needs both, because revenue might be a decade of clinical trials away.

“Deep tech sometimes requires time, but may require… less capital, right? It may be not capital intensive, but it may be time intensive. You need capital that is willing to wait.”

There’s also a guardrail on who gets in. Projects must be at “TRL 4 and above” — a NASA-derived scale where 1 is a napkin sketch and 9 is a finished product flying in the field. TRL 4 means the idea has been proven to work in a lab, just not yet in the real world. So ANRF isn’t funding pure blue-sky dreams here; it’s funding things that already half-work and need a push to commercial scale.

Where the government deliberately steps back

One detail that stands out: when it comes to the investments, the government doesn’t pick winners. RDI uses a fund-of-funds structure — it gives money to professional fund managers who run independent investment committees, and those committees decide who gets funded.

“The actual investment decisions are made by investment committees and government is not part of that.”

The honest gaps

The interview is unusually candid about what isn’t working yet. Monitoring — keeping track of whether funded projects are actually progressing — is, in his own words, “still work in progress.” He notes you can’t monitor a med-tech project and an AI project the same way; each needs domain experts who actually understand the field.

On AI specifically — the question of whether India is losing the AI race — he admits this first batch has no core AI investments. Plenty of applied AI (AI for health, AI for space data), but nothing building foundational AI itself. He says they’d welcome those proposals; they just haven’t come through yet.

And the unfinished agenda is broader than money. The hard parts, he argues, aren’t financial — they’re the connective tissue around the science: how does an early innovation become a government purchase order? How does it clear medical regulation? How do you open export markets? And how do you pull in philanthropy and corporate social responsibility money at scale to multiply the government’s contribution?

“Creating financing is just one part… How does early stage innovation lead to procurement, for example? And how does it interact with the regulatory environment?”

Key Takeaways

ANRF is a statutory body (permanently established by law), not a one-off scheme. Inside it sit multiple funds for different purposes. The Aadhaar/UIDAI relationship is the structural model.
It does two non-overlapping things: grants (no payback, to non-profits/academia) and investments (loans/convertibles, to private for-profit companies only).
The ₹1.5-lakh-crore figure is the RDI scheme, announced November 2025, run by fund managers TDB and BIRAC.
First cohort: 22 companies including Tejas Networks, Ather Energy, ideaForge, Agnikul, GalaxEye, E-Plane — spanning 6G, 3D-printed rockets, earth-observation satellites, drones, healthtech.
Ticket sizes: ₹50–250 crore. RDI funds a maximum of 50% of any project.
“Patient capital” is structured to avoid diluting founder ownership, and is matched to the kind of waiting a project needs — time-intensive vs. capital-intensive vs. both.
Funded projects must be at TRL 4+ — proven in a lab, not yet in the field. This is scale-up money, not pure blue-sky.
Investment decisions use a fund-of-funds model; independent professional committees choose, not the government.
Eligible companies must be management-controlled by resident Indians, but can partner with foreign entities and acquire technology from abroad.
Self-admitted gaps: weak monitoring infrastructure, zero core-AI funding in this cohort, and unsolved links between innovation and government procurement / regulation.

Claude’s Take

This is a corporate-relations interview, not journalism. Govindraj Ethiraj is a sharp interviewer, but the format is an executive explaining his own program — there’s no skeptic in the room, no funded founder, no critic of past Indian R&D schemes (and India has a graveyard of those). Take the optimism with that in mind.

What earns it a 6 rather than lower is the CEO’s genuine candor. He volunteers the weak spots — monitoring is half-built, there’s no core-AI investment yet, procurement and regulation are unsolved — without being cornered into it. That’s rare in a government-program puff piece, and it makes the structural explanation (grants vs. investments, the fund-of-funds firewall, the precise meaning of “patient capital”) trustworthy and clear. The Aadhaar/UIDAI analogy alone is worth the watch for understanding how the thing is wired.

What it lacks is any way to judge whether ₹1.5 lakh crore is a lot, whether the 22 companies were well-chosen, or whether this avoids the fate of past Indian science funds. Those are exactly the questions an adversarial interview would press. So: a useful, honest primer on the architecture of India’s new R&D bet — and silent on whether the bet will pay off.

India’s ₹1.5 Lakh Crore DARPA Moment For Deep Tech | Govindraj Ethiraj | The Core Report