India's deep tech funding gap has moved rather than closed. Pre-seed and seed are crowded (pretty much every fund house on this panel's radar is raising an early-stage deep tech fund) while the $30–50 million round still leans on foreign capital, corporates and family offices. The panel's fix is not more funds; it is founders trading a technology narrative for a business one, because India's late-stage money is already sitting here, waiting for evidence on a balance sheet.
44:24
The panel's shared claim is that India's deep tech funding problem has moved rather than eased. Karthik of Yali Capital says pretty much every fund house he knows is now raising an early-stage deep tech fund, so relative to five years ago pre-seed and seed are no longer the constraint; the pain starts at $30–50 million, where the ecosystem still leans on foreign capital, corporates and family offices. Manish of pi Ventures argues the block is narrative, not liquidity. Late-stage funds in India are sitting on capital and will move the moment a founder stops telling a technology story and starts telling a business one with evidence on the balance sheet, helped by defence turning buyer as it did in the US and Israel. Rishab of PremjiInvest names the structural mismatch: venture was built to price a step change in risk at every round, deep tech rounds often deliver no such step, so investors must specialise and founders should think in rupees rather than dollars. Ashish of Kotak counters that beyond Series B capital genuinely exists in India for a working product: Series A and B is what you build, C and D is financial engineering. What nobody defends is the state: grants that arrive two years late, and a regulator that took two years to move one paper on a drug whose development already runs fifteen. The stakes are that a decade of Indian deep tech gets decided less by the science than by whether founders learn to sell a business, and whether domestic late-stage money shows up before foreign money changes its mind.
Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.
How Indian deep tech capital went from absent to crowded at the early stage, and stalled at the next one.
Karthik's blunt read is that pretty much every fund house he knows is raising a deep tech fund, and almost all of them are raising an early-stage one. Compared with five years ago, he says, early-stage deep tech is simply not a problem any more and will keep improving year over year. The caveat he flags himself is definitional: what counts as deep tech may drift as the money chases the label.
Manish puts the friction precisely at the $30–40–50 million cheque, and says the cause is a transition founders have not made. Early on the story is all technology and innovation, and founders keep speaking that language long after the round requires a business story backed by evidence. His claim is that late-stage funds already exist in India and are sitting on capital. Change the narrative and those pools unlock, which is why he thinks the dependence on foreign capital fades with time rather than with new fund structures.
Manish says his frustration is with the government, which opens large pockets of money and then makes everyone chase bureaucracy (ten visits, a thousand forms) and cites a portfolio company whose grant arrived two years later, by which time it no longer needed the cheque. Ashish points the same finger at regulators: one company waited two years for a single paper to move at the drug regulator, on a molecule whose development already runs fifteen years, and another filed a first-in-class application that the office had no precedent for. Both agree the intent is right and the gap sits between intent and the bureaucratic layer.
pi Ventures raised an AI fund in 2016 having failed to find AI companies in India, then wrote one of that fund's last cheques into 3D-printed rocket engines because a founder refused to be ignored, and material science, not software, shaped Fund II. Manish's summary is that investors think they are knowledgeable but are actually gamblers betting on a fundamental: some technology will solve today's problems 10x better, and it will be disruptive rather than incremental. Karthik makes the same admission prospectively, expecting something to walk into the boardroom that forces a rejig of the thesis.
Rishab's structural point is that venture capital grew up around technology investing, where each round retired a discrete chunk of risk and capital was available for each step. Deep tech takes far longer, and many funding requirements arrive without a meaningful step change in risk at all, which is exactly why the early-to-Series-B stretch feels like friction. The remedy cuts both ways: investors have to specialise, and founders have to identify early on what their next real step change in risk actually is.
Karthik reframes dependence on overseas money as an underappreciated benefit: a foreign investor at Series A-plus brings expectations of execution that most Indian startups have never experienced. His line is that the ecosystem aspires to Silicon Valley valuations without having internalised Silicon Valley execution. Manish agrees the discipline is real but says it is customer-driven rather than investor-driven, because a global customer will not keep buying unless you deliver at their pace, and that domestic exceptions who execute well will eventually reset the culture as role models.
Asked whether a semiconductor startup should run consulting on the side for optionality, Karthik's answer is don't even think about it: getting to financial viability before the money runs out or someone disrupts you already puts you on a knife edge, and diverting energy for incremental capital costs the speed you need to move from FPGA validation to test chip to commercialisation. None of those milestones produce revenue, he notes, but all of them should produce value multiples. Manish dissents mildly: services money is easy to get in India and makes a fine profitable business, just not a venture-funded one, so decide which of the two you are building first.
Ashish dates the shift to the last six or seven years: grants and government schemes gave scientists cover to attempt companies, and researchers who left in the late 2000s, when none of his peers imagined founding anything, came back from 2016-17 to build. On the other side, pharma companies built over a decade or more of generics manufacturing are sitting on cash they do not know what to do with, and have started investing into early clinical assets, something he calls unheard of. His own book now carries nine clinical-asset companies.
Every figure below was said on air, with timestamps included and caveats kept.
As investors we think we are knowledgeable, but we are actually gamblers.
— Manish, pi Ventures ▶ 10:17
We aspire to Silicon Valley valuations, but we have not really internalised what Silicon Valley execution really is. If you've not experienced it, it is something else altogether.
— Karthik, Yali Capital ▶ 30:56
Series A and B is all about what you build. Series C, D and everything beyond that is financial engineering.
— Ashish, Kotak Alternate Assets ▶ 33:50
People are changing their boards, but they have to change their frameworks. If you're a SaaS investor looking for product-market fit and you talk in that framework to a deep tech company, it will be a miss.
— Manish, pi Ventures ▶ 38:43
Don't even think about it. The effort it takes to get a semiconductor startup to financial viability, before the money runs out and before someone comes and disrupts you — you're already on the knife edge.
— Karthik, Yali Capital ▶ 40:55
How pi Ventures went from an AI-only fund nobody could deploy to a ₹700 crore deep tech book of mostly physical innovation, and the founder who forced the change.
The grant that arrived after it was needed, the drug regulator that added two years to a fifteen-year timeline, and both investors agreeing intent is not the problem.
The clearest statement of the thesis: early money is abundant, Series B onward still depends on foreign capital, and India lacks funds that can carry a company to IPO.
The transition from technology story to business story, why late-stage Indian capital is already sitting there, and defence arriving as the buyer that changes balance sheets.
A live audience question splits the panel, knife-edge focus against a perfectly good non-venture business, with the semiconductor value milestones spelled out.
Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.