Institutions produce conviction

An institute-backed deep-tech VC? This is how ecosystems are built.

I had the privilege of being at the launch of India's first institute-backed deep-tech venture fund, by the Society for Innovation and Entrepreneurship (SINE) at IIT Bombay. It did not feel like just another fund launch. It felt like a blueprint for how India can seriously back science-led entrepreneurship.

An ecosystem, not just an incubator

The first thing that struck me was how SINE thinks of itself. Not as an incubator that houses startups for a while, but as an ecosystem.

From Atomberg to ideaForge, a growing list of deep-tech companies has come out of this environment. That list did not happen by accident. It is what patient institution-building looks like when you give it enough time.

Deep tech needs deep patience

India still has a shortage of long-term, risk-tolerant capital for hard problems. Software startups can show traction in months. A new battery chemistry, a drone platform or a semiconductor design can take a decade before it earns its first real revenue.

Most venture funds are not built for that clock. Their fund cycles, their return expectations and their portfolio math all push towards speed.

Institute-backed capital aligns research timelines with venture timelines.

That alignment is what makes this model interesting. A fund that sits next to the lab understands why a prototype takes three years. It can tell the difference between a technical setback and a dead end. It can back founders through the valley that kills most hardware companies.

Why deep tech breaks the usual venture model

It is worth being precise about why hard-science companies struggle with conventional venture capital. The difference is not just time. It is the shape of the risk.

A software startup mostly carries market risk: will customers want this? A deep-tech startup carries technical risk first: can this be built at all, at the right cost and at scale? Only after that does market risk arrive. Most investors are trained to evaluate the second kind of risk, not the first.

Deep-tech companies also need capital in larger, lumpier amounts. Labs, pilot lines, certification and manufacturing scale-up cost far more than cloud credits and a sales team. And the exit paths are different. Buyers are often industrial companies or governments, not other technology platforms.

Put together, that means deep tech needs investors who can judge the science, tolerate long timelines and stay engaged through manufacturing. Very few funds are set up to do all three.

Talent density is leverage

Premier institutes are not just idea factories. They are living labs and test beds.

When founders, faculty and students collide in the same physical space, ideas get stress-tested early. A professor spots the flaw in the physics. A student builds the first prototype over a weekend. A senior founder down the corridor shares what went wrong with their first factory. Business models mature faster when the feedback loop is this short.

This is something no amount of capital alone can buy.

The courage to absorb failure

What made the evening even more grounding was hearing Dr K. Radhakrishnan, former Chairman of the Indian Space Research Organisation, speak candidly about ISRO's tryst with failure.

His point was simple and powerful. World-class capability is not built by avoiding failure. It is built by institutions with the courage to absorb it, learn from it, and keep going. Over decades, that courage compounds.

ISRO is a reminder that India can build world-class deep technology on its own terms, when patience, talent and institutional backing come together.

The lesson for corporates

There is a lesson here for large companies too, not only for universities and investors. Many corporates want to access deep-tech innovation, through venture arms, partnerships or open innovation programmes. Many are disappointed by the results.

The reason is usually a mismatch of clocks. Corporate budgets run annually, executives rotate every few years, and pilot projects are judged on quarterly milestones. Deep-tech partners need a sponsor who will still be there when the third prototype finally works.

The companies that do this well treat deep tech as a long-term portfolio, not a series of experiments. They protect a small budget from annual cuts, give one senior leader lasting ownership, and measure progress by technical milestones rather than early revenue.

A model worth replicating

If India is serious about deep-tech leadership, and not just unicorn headlines, this model deserves to be replicated. Thoughtfully, patiently and at scale. Other institutes have the talent. Many have the research. What they need is capital that thinks in decades, and partners in industry willing to be first customers.

Institutions don't just produce talent. They can produce conviction. In deep tech, conviction is the scarcest resource of all.