Qualcomm Ventures eyes India startups to build low-cost A…

Qualcomm Ventures’ new $150 million fund will back founders building lower-cost AI applications, robotics and edge technologies

Chipmaker Qualcomm’s venture capital arm is looking to back Indian startups building artificial intelligence (AI) technologies affordable enough to scale globally through its new $150 million India-focused fund, according to a top executive at the firm. With its second India-focused fund, Qualcomm Ventures is looking for founders who are reducing the cost of deploying AI applications, an area where the firm believes Indian startups have a natural advantage.

"Looking at the Indian startup ecosystem, companies here are in the unique position to make AI affordable globally," Rama Bethmangalkar, managing director at Qualcomm Ventures India, told Mint. “With the population and purchasing power that the country has, companies can't afford to make solutions which are not designed with the Indian price point in mind.” Qualcomm's chief executive Cristiano Amon had announced the AI-focused $150 million fund at the India AI Impact Summit held in Delhi last month.

With the new fund, the firm plans to invest across deeptech, AI applications spanning both B2B and B2C markets, and enterprise technology. San Diego-based Qualcomm Ventures, at least in India, is particularly focused on companies built around artificial intelligence from the outset rather than startups adding AI features to existing products to keep up with the surge of interest in the technology.

“Fundamentally, founders who are pushing the envelope on difficult problem statements and are looking to solve them from an Indian perspective are the kind of people that excite us,” Bethmangalkar said. One area the firm is especially bullish on is physical AI, including robotics systems that can automate industrial workflows, he said. It's an area where companies and investors are increasingly looking to reduce labour while simultaneously ensuring minimal downtime.

Premji Invest recently led a $450 million round in AI robotics company Rhoda AI with participation from Khosla Ventures, Temasek Holdings and Silicon Valley billionaire John Doerr, valuing the company at $1.7 billion. Homegrown robotics company CynLr last month announced a new tech suite which lets robots learn on the fly, much like a human child, instead of requiring reinforcement learning. Hyundai, through its subsidiary Boston Dynamics, has pledged to build and deploy 30,000 units annually by 2028 humanoid robots to place in its manufacturing plants across the US.

Another focus area for Qualcomm Ventures is companies developing AI for edge devices such as phones, earphones, smart glasses, smart rings, bracelets and pendants. “There are going to be some very exciting products in the next couple of years is our belief,” Bethmangalkar said. “But for it augment users and create richer experiences, phones will become a companion inferencing engine for that device.” “Interest in deep tech and AI in India is steadily increasing, but it will not follow the same rapid funding cycles seen in other startup sectors,” said Amit Nawka, partner, deals and startups leader at PwC India.

“The space involves longer development cycles and the added uncertainty of fast-evolving technologies. As a result, investors are often placing a wider set of bets across the ecosystem, recognizing that it will take time to see which technologies and business models ultimately prevail.” India's venture capital ecosystem rebounded in 2024, according to a Bain & Co report released last year, with investments growing 1.4x to $13.7 billion, up from $9.6 billion even as funding in the larger Asia-Pacific region remained flat.

Artificial intelligence startups attracted $1.2 billion in venture capital last year, with the sector’s share of overall funding rising to 12% from under 5% in 2020, according to a report from the India Deep Tech Alliance, whose members include Accel, Blume Ventures, 3one4 Capital, Chiratae Ventures, Nvidia, Premji Invest, and Qualcomm Ventures. The fund has not set a target for the number of investments it will make.

Qualcomm Ventures expects its cheque sizes to average around $5 million, with a preference for investing from Series A onwards, once a startup has demonstrated either a minimum viable product or product-market fit. It is also open to investing earlier or later depending on the opportunity. Qualcomm has entered Indian startups at Series B and C in the past. “Pre-seed is fine if its something very strategic and where we think we can bring some value right off the bat,” said Bethmangalkar.

The firm exceeded the size of its maiden India fund from 2015, which was also set at $150 million, though Bethmangalkar declined to say by how much, saying only that it was “exceeded by a significant mile.” Unlike most venture funds, Qualcomm Ventures invests directly from Qualcomm’s balance sheet and operates with a single limited partner structure, allowing it to stay invested in companies longer without the pressure of a fixed exit timeline.

Since it began investing in India in 2008 with MapMyIndia, the firm has backed around 40 companies in the country. "We've had that discipline to return capital back to corporate. As a result, we have a healthy pipeline of investments and the cycle continues because we're able to balance both strategic and financial objectives," Bethmangalkar said. Companies in its portfolio that have gone public include MapMyIndia, IdeaForge, Capillary Technologies and Shadowfax.

Companies expected to go public include audio device maker BoAt, FabHotels and Tonbo Imaging.

Summary

This report covers the latest developments in artificial intelligence. The information presented highlights key changes and updates that are relevant to those following this topic.

---