While visiting the Kerala Start-up Mission conference, VMware's VP-Technology and managing director Ramkumar Narayanan, also the chairperson of Nasscom's Deep Tech Start-Up Council, was impressed with the use cases presented by deep tech companies. Though he has been tracking this sector for a while, he was amazed at how these deep tech start-ups are solving real-world problems, and he believes it augurs well for the future.

Narayanan discusses the role Nasscom plays in the deep tech sector in an exclusive interaction with Outlook Start-Up. He also talks about the challenges faced by founders and the possible solutions to pave way for thousands of deep tech start-ups by 2030.

Edited excerpts:

Deep tech start-ups have a longer gestation period, varying from six to eight years before they have a marketable product. At the same time, they are reliant on investor capital and institutional investors who expect faster returns within three to five years. How can industry bodies like Nasscom bridge this gap? 

Nasscom focuses on the requirements and talent strategy of deep tech start-ups, supporting them with market access domestically and internationally and helping in policy creation. Typically, early-stage funding primarily comes from government or institutional grants. However, these are meagre for deep tech start-ups. 

Recently, some space tech start-ups bagged a Rs 20 lakh grant. A satellite tech company is building their first proof of concept for around Rs 3 crore, so this grant won't get them very far. Comparatively, deep tech companies in the US have access to grants in the range of $20,000, depending on where they are getting it from. 

The government is trying to address this by bringing disparate pools of money under a common pool in a fund-of-funds model within the National Research Fund. Modalities are being worked out on how founders can get access to this Rs 50,000 crore capital, how it will work for their projects, etc. 

Since not everything can come from the government, the other source for grants is from institutions like incubators and accelerators. The last capital source that founders tap is the friends and family model since their timetable is more flexible than institutional investors. 

Fortunately, traditional VCs dabbling into deep tech understand that this sector follows a different capital timetable. While they might look at annual recurring revenue (ARR) for B2B or SaaS start-ups, they realise this may not apply to deep tech start-ups given their lengthy gestation period. Founders in the sector are also looking at LPs like family offices, who are not looking for the traditional five-to-eight-year returns. 

As per various reports, around 10 per cent of the 28,000 tech start-ups are focused on deep tech. Will these initiatives see a sharp increase in this number?  

Historically, the tech ecosystem propagates nonlinearly with the early pioneers struggling the most. Currently, there are around 3400 deep tech start-ups, which we expect will rise to 10,000 by 2030 as more companies get funded. 

Nasscom has helped 1000 deep tech start-ups already through its cohort, which it started in 2017. This year's Nasscom Product Conclave focused on deep tech and saw great participation from founders. 

The recent Nasscom Future Forge also announced a consortium of institutional investment partners. It will provide a single window for start-ups looking for funds so they don't have to go from VC to VC. Even Nasscom's flagship 10k program is morphing to support deep tech. 

What identification and filtration process does Nasscom follow for including start-ups in this deep tech cohort? 

When we started our first cohort in 2017, our selection rate was about 6 per cent of the application pool. These companies were focused on health, cybersecurity and drones; some are thriving in India and globally. 

Deep tech is driven by research and longevity. Our touchstone is the research that the company is investing in because it takes almost seven years before they can launch a product. We are seeing some great start-ups focusing on health, agriculture, energy, climate, tech and mobility. 

We are also helping these companies with their go-to-market strategies, which include the product, supply chain piece and regulatory environments. I am not a big fan of boiling the ocean and instead prefer to pick fewer but crucial problems in the sector as the Nasscom chair for the deep tech council.  

While there is a lot of buzz around genAI and how it can transform the overall economy, no one seems to have any clarity on when it will show higher gains in terms of productivity and growth. What is your prediction?

In the early 2000s there were almost 300 web search providers. But as the economics kicked in, the realisation set in that web search for the larger public is an extremely expensive game. 

Now, you can't put the genAI genie back in the bottle. However, not everybody can put the kind of financial commitment that Microsoft has with its $10 billion investment in the technology and its initiatives to provide infrastructure for that use. 

So, there will soon be only a few players who can provide genAI experience, like OpenAI or Google Bard. These are the usual suspects because they have the muscle and money to provide this kind of support. These companies will soon focus on vertical problems, like helping enterprises reduce customer support costs and help their agents become more productive. 

This will help companies support more customers in a shorter period or eventually open it up for customers to self-serve. Enterprises can pick and train any open-source model with their data without needing Microsoft's gargantuan clusters and servers.

This differentiation will occur where there will be large players and smaller ones that are purpose-built for solving select problems in organisations. 

While there is a lot of worry about the displacement of professions due to gen AI, is there any clarity about the kind of jobs that will be lost or how organisations can redirect these employees into other areas?

When banks were automated years ago, there was a lot of pushback from the banking unions that jobs would be lost. I don't know how many jobs were really lost, but the banks computerised and made their customers and workforce more productive.

Similarly, when it comes to genAI, certain kinds of jobs will not be needed, but mass layoffs are unlikely to occur overnight. Over time, the workforce will need to adjust to newer needs to become more productive, creating new opportunities. 

For instance, a software developer currently spends a lot of time not just coding or debugging but also adhering to organisational norms. So, genAI can become their co-pilot that can care for coding standards and templates or testing frameworks. It will make the developer more productive and get outcomes faster. 

The way gen AI will probably play out is that it will create a new class of jobs, especially within a function. That will help businesses to take on more work or customers and become more productive.

At the same time, we can't trust genAI in its current form because of hallucinations and other issues. So, a human presence in the middle is still required. 

• 
• 
• 
•