Bharat Semi Raises $24M Seed to Build the Sensor Layer for Physical AI — and What Asia's Deep-Tech Founders Should Learn About Pitching It
Bharat Semi, a New Delhi-based semiconductor startup, announced a US$24 million seed round on 16 November 2026, led by Info Edge Group, South Park Commons and Indus Bridge Ventures, to build what the company describes as "the sensor layer for Physical AI" — image sensors that combine sensing, processing and artificial intelligence on a single chip. The company is targeting consumer electronics, surveillance, robotics, automotive and aerospace applications. For founders across India, China, Singapore and the wider region, the round is a useful case study in two things at once: how deep-tech teams raise serious capital for complex silicon, and how they explain that complexity to investors who will never read a datasheet.
Want your next investor or sales pitch to land more clearly? Explore Jason Lim's Presentation Skills and Speaking Program — "How to SELL. WIN the Deal." It starts with a free consultation offer (by application).
What the US$24M Is For
The US$24 million seed was announced on 16 November 2026 and led by Info Edge Group, South Park Commons and Indus Bridge Ventures. Other participants included Valour Capital, Micromax and Gaingels, along with angel investors including Mukul Agarwal and Karthik Sundar Iyer. According to IncBusiness, the funds will be used to grow the engineering team and take the company's sensors to market.
Writing for Entrackr, Shashank Pathak noted that seed rounds of US$24 million or more remained relatively uncommon among Indian-origin startups in 2026 — a reminder of how unusual it still is for a hardware and silicon company at this stage to command this kind of cheque in the region.
The Founders and the Technical Bet
Bharat Semi (bharatsemi.ai — note the .ai domain, not .com) has offices in New Delhi, Bengaluru and San Francisco. It was co-founded by Vrinda Kapoor, Co-founder & CEO, and Vinayak Dalmia, Co-founder & Managing Director (listed on the company's team page). The company's CTO is Boyd Fowler, formerly CTO at OmniVision.
The technical pitch, as the company tells it: today's AI systems typically move raw sensor data across multiple chips before anything intelligent happens with it. Bharat Semi says it is building image sensors that combine sensing, processing and AI on one chip — what it calls the sensor layer for Physical AI, meaning machines that perceive and act in the physical world. The company claims (unverified) that its approach can deliver up to 99 per cent less data movement between components and up to 10x lower system costs compared with conventional multi-chip modules.
Per an IncBusiness timeline (cited by name only; its page is currently unavailable), the company planned a fab in 2024 with a US Space Force technology partnership, and signed an MoU with Kaynes Semicon in September 2025.
What the Founders and Investors Said
Business Standard (Peerzada Abrar, 16 November 2026) quoted the founders and their lead investor on why now:
"In the near future, growth in the sensor market will be driven by capturing data for machines. This data may never be seen by humans, but it will drive AI at the edge… AI algorithms and hardware have improved enough to make Physical AI a reality, and we intend to define this category." — Vinayak Dalmia, Co-founder & Managing Director, Bharat Semi
"Every sensor today streams data to intelligent systems located remotely. Intelligence at the right place is a meaningful paradigm shift… there is an opportunity for India to build a globally relevant product company in this space." — Vrinda Kapoor, Co-founder & CEO, Bharat Semi
"Bharat Semi is tackling a significant limitation in Physical AI at its origin, within the silicon itself." — Vibhore Sharma, Redstart, Info Edge deeptech
Related News
- Jason Lim: From Rock Bottom to Sales Coach Behind Millions in Founder Deals
- He Built a Startup at BAJA SAEINDIA. Twenty Years Later, He's the Keynote Speaker.
- Botsync Raises US$11 Million Led by IvyCap Ventures: How Singapore's 'Physical AI' Startup Tells Its Story Across Borders
Our Analysis: What This Means for Asia's Founders
The following section is our analysis and opinion, not reported fact.
The most interesting line in the coverage is Dalmia's "we intend to define this category." That is a category claim, not a market position: no independent evidence yet shows that single-chip sensing-plus-AI is a category anyone else recognises, and the company's performance claims (up to 99 per cent less data movement, up to 10x lower system costs) remain company claims, unverified. Seed-stage hardware promises routinely take years to prove, and many do not survive contact with volume manufacturing.
Still, the shape of the round matters for Asia's founders. Info Edge — best known for consumer internet bets — leading a silicon seed suggests regional capital is willing to underwrite deep hardware if the story is framed as a market shift ("machines, not humans, will consume sensor data") rather than a component roadmap. If integrated sensing-and-compute does become a real category, India and China are both plausible contenders for vertically integrated sensor/AI stacks, each with large domestic demand and growing design talent. Singapore, without a comparable fab base, is more credibly positioned as a partnership and capital hub for such efforts — a connector role rather than a silicon role. That is a hedge, not a forecast; the honest takeaway is that capital is arriving before the technology is proven, and the teams that can narrate that gap credibly will raise better than those that only show datasheets.
How Can Asia Startups Compete on the Global Stage?
"More than half of the 2021 Fortune 500 companies offer in-house Toastmasters clubs to help build and improve their employees' communication, leadership, and public speaking skills" — Toastmasters International Media Center, 26 April 2022. The same source notes that nineteen of the top 20 Fortune 500 companies — and 85 of the top 100 — have at least one active Toastmasters club.
The world's largest companies treat speaking as infrastructure, not innate talent, and that is the transferable lesson here. Kapoor and Dalmia's quotes work because they translate a silicon architecture into a business argument: name the shift (data for machines, not humans), name the constraint (intelligence located remotely), name the ambition (a globally relevant product company from India). Asia's deep-tech founders competing globally need the same discipline — the science does not translate itself, and investors fund the version of the story they can repeat to their own committees.
If you are preparing an investor or customer pitch and want it to land more clearly, explore Jason Lim's Presentation Skills and Speaking Program — "How to SELL. WIN the Deal." It begins with a free consultation offer (by application). Results depend on your own effort and market conditions; no specific outcomes are guaranteed.
Ad: DreamakerAI, an AI CRM and business operating system.
Disclosure: Bharat Semi is not affiliated with and did not endorse the program mentioned in this article.
Sources
- IncBusiness — use of funds and company timeline (cited by name; its page is currently unavailable)
- Business Standard — Peerzada Abrar, 16 November 2026
- Entrackr — Shashank Pathak, 16 November 2026
- bharatsemi.ai — company website
- Toastmasters International Media Center — 26 April 2022
