Manish Singhal has two video technology patents to his name and helped build products at companies such as Motorola and Sling Media before becoming an entrepreneur and investor. No wonder, then, that the venture capital fund Singhal co-launched last year—pi Ventures—focuses on tech startups, particularly those working on artificial intelligence and machine learning.
The VC fund aims to raise $30 million; it has already attracted investors such as International Finance Corporation, Accel Partners, Flipkart co-founder Binny Bansal and prominent angel investor TV Mohandas Pai. The fund has backed four startups and will invest in a total of 18-20 startups. In an interaction with VCCircle, Singhal speaks about AI startups in India and the fund’s strategy. Excerpts:
How many AI startups does India have and how are they evolving?
According to a Zinnov Consulting study in October last year, India is the third-largest cluster after the US and UK with around 200 AI startups. China does not figure in the list. I feel the China data is not easily available. If China is probably ahead of us, then we are the fourth-largest [cluster].
We [pi Ventures] have done four investments in the last one year. We might have come across 200 companies and have met around 100 companies. The Zinnov report is eight months old and the number of AI startups could be more than what it said as we ourselves have come across over 200 startups. But, obviously, when you look at it from the lens of investment, the pie reduces.
There is a lot happening in this area. It is unravelling dramatically. Growth will be nonlinear in this space.
What is triggering this growth?
Data has become a very relevant point. Whatever you do produces data. Businesses are producing data; data is all around us. Products of the future would be [those] that use these data more intelligently than others. With computation power and networks available, you can afford to run artificial intelligence solutions and derive data and real-time intelligence.
Our theory is that for data to play a big role, AI must play a bigger role. More companies seem to be adopting it to make a difference to their customers. The next wave will come from AI-powered products.
What is your approach when you invest in companies?
We look for companies that use very deep technology, but which can be applied. We are extremely particular about whether they have a business case that they are solving. If they have just technology, we may not get excited. But if they have a business case that they are solving with world-class technology, then we get excited.
When do you think AI startups in India would achieve a certain size?
AI is a means to an end. All these companies are solving real problems. We haven’t seen a big winner yet. Maybe in another two to three years, we should see some global names coming out of India.
If you look at the applied side of artificial intelligence, the whole world is very young. You can hardly name any company. Earlier, a lot of AI platform companies came up and that led to a lot of acquisitions in this space. India is a great case for applied AI. There are so many problems to be solved. [There is] so much chaos in many sectors. You can actually leapfrog with automated solutions.
What is your ticket size and at what stage do you enter?
We are seed investors and we go in as early as possible; at times at idea level also. We have not [invested] at the idea level yet, but we are hoping to do so if the idea is disruptive. We have been a hands-on company and product builders. In certain companies, we have gone to the coding level also.
Typically, we get in where you have basic technology and proof that this technology works; it may not have gone to the market yet. In those stages, what we see is that companies would typically need $100,000-200,000 and a median of around $500,000. That is our sweet spot. What we look for is
essentially, as I said, if it solves a problem in a differentiated way.
AI is a fledgling area and is unpredictable. Do you think the risk is higher in this space?
These are intellectual property-led businesses, so there will be a residual value if the companies don’t do well. Founders get time to create their business. It is not ‘throw the money and see the magic’.
There is a lot of acquisition play here. The worldwide average of acquisitions in the AI space is four to five years. We would potentially see acquisitions and that would give us exits much sooner. We have planned for a longer horizon, seven plus two years. That is a good time for a seed investor to stay invested.
Like this interview? Sign up for our daily newsletter to get our top reports.