Why Deep Tech Founders Leave India

By Priyanshu Jangra Founder/ CEO · Tier 1 · 2026-07-24

Anagha Rajesh spent two years building BioCompute in Bengaluru — DNA data storage, encoding information into molecules instead of magnetic platters, because data centres are running out of land and power. Over ₹5 crore raised. A lab. Thousands of experiments. A working prototype, and as far as public reporting goes, the first such lab in India.

Then, in mid-2026, she shut it down. Let the team go. Listed the equipment for sale on LinkedIn. Moved to San Francisco.

Her explanation travelled further than her science did. In SF, people asked what she needed to finish the work. At home, the first question was when revenue starts.

Indian startup Twitter knew what to do with that. Brain drain. Cowardly capital. Another one gone.

That reading is wrong, and the wrongness costs founders money.

Capital has a clock

A fund that passes on your ten-year bet is usually not being timid. It is being compliant.

A venture fund is a contract. Limited partners hand a manager money and expect it back, with returns, inside a defined window — typically ten years. A fund deploying in year six needs liquidity in four. Your DNA storage chip will not provide it. Neither will your semiconductor or your battery chemistry.

So the investor across the table isn't rejecting your science. They're telling you, badly, that their clock and your clock don't match. You cannot shame a fund out of its own mandate. You can only find capital with a different clock.

That reframes the question. Not "does India believe in deep tech," but "does India have a capital pool with a fifteen-year clock, and do you know how to reach it."

Until recently the honest answer was mostly no.

What changed while everyone was arguing

In July 2025 the Union Cabinet approved the Research, Development and Innovation Scheme; it launched that November. ₹1 lakh crore over six years, housed under the Anusandhan National Research Foundation, with the Technology Development Board and BIRAC as second-level fund managers.

I'd have ignored it too. Government scheme announcements are a genre. But the terms are strange in a good way.

Twelve to fifteen years at three to four percent is neither a grant nor venture capital. It's built for the phase where deep tech companies actually die.

**That phase is the second valley of death: the gap between a working prototype and a manufacturable product.**Grants and angel cheques usually cover reaching a prototype. The next stretch costs an order of magnitude more, involves manufacturing and regulatory work, and produces no revenue while it happens.

Is it moving? Partly. The first call for proposals opened in February 2026, drawing a reported 124 proposals worth over ₹25,000 crore by April. First agreements were signed on 13 May 2026 with five companies including Dhruva Space and Noccarc Robotics. By July, TDB was targeting roughly ₹500 crore disbursed by September.

Five hundred crore against a one lakh crore headline is a rounding error, and deployment speed will stay an open question for years. But something exists now that didn't in 2021, and plenty of founders are still deciding on 2021 information.

The line in the eligibility criteria

Reported eligibility requires an entity registered in India and controlled by resident Indian citizens.

Read that with a founder's eyes. The instrument in this country whose duration best matches a deep tech company's actual life is one you stop qualifying for the moment you flip your holding company to Delaware.

That's not an argument against leaving. It's an argument against leaving casually, or early, or for optionality — which is how most flips actually happen. A seed-stage founder redomiciles because someone said US investors prefer it, and forfeits domestic capital for a US round still three years out.

When leaving is the right call

Three conditions make relocation defensible. Meet two, and go.

Your buyers are elsewhere. If seven of your first ten customers are American hyperscalers or defence primes, proximity removes real commercial risk. This is the strongest argument and it has nothing to do with investors.

Your next round exceeds domestic depth. India's early stage has improved enormously; the corridor above Series B for pre-revenue hardware has not. Founders rarely leave over ₹5 crore. They leave because they can't see ₹50 crore.

The person you need isn't here. Someone who has shipped exactly this before. Real constraint — but note that moving a company is an expensive way to hire one person.

The underused middle path: keep the Indian entity and the R&D, open a US subsidiary for sales, preserve eligibility on both sides. Less drama, more paperwork. Talk to a cross-border structuring lawyer first, because unwinding a flip later is slow and expensive.

What to actually do

Map milestones to Technology Readiness Levels rather than revenue projections — that's the language patient-capital instruments are written in. Know your bill of materials at scale, not lab scale. Name your first ten customers. Be able to answer what happens if this takes twice as long, because founders lose rooms there far more often than on the science.

Then match instrument to stage: non-dilutive and concessional money through the valley, when your valuation is lowest and your capital need longest. Equity later, once something is manufacturable. Invert that order and you sell most of a company whose value gets created a decade after you sold it.

The part I keep coming back to

Some founders will still leave, and some should. Companies aren't national flags.

But "I couldn't raise in India" and "I didn't know what to apply for in India" are different sentences, and more people are saying the first while meaning the second than anyone wants to admit.

The useful question was never whether India believes in you. It's whether you've matched your capital to your clock.

If you're building something that takes years before it earns anything, Pitchsap helps founders map the roadmap stage by stage, find mentors who've raised for hardware and biotech, and get investor-ready before the first conversation rather than during it.