Indian Founders Now Ready to Renounce Quick Returns for Deep-Tech Innovation: STPI DG

In fact, the largest start-up ecosystem in India is in Tier II and Tier III cities. STPI is one of the biggest contributors to this ecosystem. We run multiple schemes. One of them is called the Centre of Entrepreneurship, under which we incubate start-ups and provide a small seed fund ranging from ₹5 lakh to ₹15 lakh. This is not an investment in exchange for ownership. We do not take investment equity. We only take incubation equity, typically between one and three%, and the start-up remains incubated with us for three to four years.

During this period, start-ups can participate in multiple programs, gain investor access, receive mentoring, and be part of incubation initiatives. All of this is provided free of cost, apart from the small incubation equity.

We also run another scheme that is purely investment-focused, called the Next Generation Incubation Scheme. Under this, we provide support of up to ₹25 lakh per start-up, where we do take investment-linked equity.

Beyond these schemes, STPI has 68 centres across the country that function as incubation spaces. Any start-up or small IT company can access these centres. Even if a founder does not want mentoring or a formal incubation program and only needs two or three plug-and-play seats to start operations, they can use any STPI centre.

We are present in cities like Darbhanga, Bhagalpur, Angul, and many other locations across the country. Start-ups can use these incubation spaces without giving up any equity and still remain connected to STPI’s broader infrastructure and ecosystem.

I would not say the sentiment has changed yet. AI is still a relatively new technology, and the most critical factor is access to talent. When a start-up is just getting started, founders cannot do everything on their own. They need two or three skilled people to work alongside them, and today, that kind of concentrated talent is still most easily available in Bengaluru.

Talent is gradually being created in Tier II and Tier III cities, but this process will take time. What we are seeing now is an interesting transition. Recently, when I was in Patna, I met two or three start-up founders who had earlier worked in Bengaluru or Hyderabad and have now returned to their hometowns.

In some cases, this move is driven by personal reasons, such as family responsibilities. One founder, for instance, returned because his father did not want to relocate, and he wanted to stay close to his family. He has now started his company from our incubation space in Patna.

This is how local talent pools begin to form. People who have gained experience in cities like Bengaluru, Hyderabad, or even abroad, come back to their hometowns. They often take a pay cut, but they are willing to take that risk because of family ties, a sense of belonging, or a desire to be part of their local culture.

Over time, these returnees help create a strong talent base. Once that happens, larger companies and new start-ups also start looking at these cities as viable locations. This is the natural way in which start-up ecosystems in Tier II and Tier III cities evolve.

One key challenge we are seeing today is a shift in the kind of start-ups being built. Earlier, many start-ups followed aggregator-type models. These were relatively quick to launch. You could build an app, acquire some customers, and start generating revenue fairly quickly, such as services promising a doctor or an ambulance within minutes.

Now, many founders are moving towards AI-driven and deep-tech models. These have a much longer gestation period. They are not low-hanging fruits.

Because of this, start-ups are now looking for more patient capital. They need larger funding amounts, as ₹25 lakh is often not sufficient to sustain a deep-tech start-up for three years. Even hiring two people and paying salaries over that period can quickly exhaust such capital.

This is the key change I am observing. Founders are more serious, are building deeper technology, and are seeking larger seed rounds, often in the range of ₹1–2 crore. They also understand that returns will take time. The focus is shifting away from quick, traditional tech models towards long-term, deep-tech innovation.

Anyone who learns AI will continue to remain relevant. There may be some smaller or routine tasks, such as drafting letters or handling basic BPO work that AI can automate through chatbots. However, someone still needs to design, build, and maintain those chatbots. In that sense, while AI replaces certain tasks, it also creates new forms of employment.

Developing deep-tech products, building vertical AI applications, and creating intelligent systems require a large number of people who understand AI. In that sense, we are moving towards a situation similar to the early days of computers. Today, we broadly have two kinds of people: those who know how to use computers and those who do not.

A similar divide may emerge with AI. Those who do not learn AI may struggle to find jobs or may be limited to more conventional roles. However, those already working in IT who upskill and reskill themselves should not face major employment challenges.

India has a very large domestic market and an even larger global opportunity. In IT and IT-enabled services, India already commands more than 50% of the global market. There is no reason we cannot achieve a similar position in AI as well. The key is to unlearn, learn, and adapt.

The same applies to deep-tech and product-led companies. These ventures have much longer gestation periods and require patient capital. As a developing country, India has traditionally favoured models that deliver returns quickly. However, as I mentioned earlier, this mindset is now beginning to change.

This shift is also why the government is placing strong emphasis on creating Centres of Excellence. A large fund of around ₹1,000 crore has been created under the Department of Science and Technology, and several Centres of Excellence have been set up across IITs. STPI has also established Centres of Excellence focused on entrepreneurship and emerging technologies.

These initiatives may not immediately produce large, breakthrough products, but they play a crucial role in building a strong talent pool. Over time, this talent can go on to create globally competitive products. For example, Zoho has built strong alternatives to widely used enterprise tools through sustained effort over many years.

Similarly, many start-ups today may be working on relatively small products, but they are contributing to talent creation. The indirect advantage of Centres of Excellence lies not just in immediate product outcomes, but in nurturing skilled professionals who can later build large, potentially AI-driven products.

That said, talent remains a key challenge. There is currently a significant gap between demand and supply, with only about 50% of the demand for AI talent being met. However, the government has launched several initiatives to address this. The Ministry of Electronics and Information Technology, for instance, plans to set up over 500 AI labs, and multiple institutions are already working in this direction.

While the talent gap is real today, it is a natural part of an evolving ecosystem. With these efforts, it is likely to narrow significantly in the coming years.

I don’t think confusion arises unless a start-up is trying to take advantage of multiple incubators at the same time. Many incubators run several funding schemes, and when start-ups attempt to access multiple schemes through different incubators, that is when confusion sets in.

Otherwise, if a start-up is straightforward, a single incubator is usually sufficient. Most incubators can provide the required number of seats, whether it is four, five, or six. It is also common industry practice for incubators to take a small equity stake, typically ranging from one to three percent. Some incubators have additional conditions. For instance, IIT incubators may require the mentor to be from their own institute, such as a professor or associate professor.

Other incubators allow start-ups to choose mentors freely but take one to three percent equity and do not charge for plug-and-play infrastructure. Some follow a different model, charging for workspace while offering greater exposure through events across the country and stronger investor connections.

Multiple incubation models are currently in place, and the right choice depends on the start-up’s stage, location, and what it is specifically looking for from an incubator. Incubators do not approach start-ups; start-ups approach incubators. Founders should carefully examine the terms and conditions and choose the incubator they are most comfortable with.

Problems and confusion arise only when greed takes over, and start-ups focus more on collecting funds than on building their product.