SF Startup Builds Wearable for Voice AI in Noise
A public beta is planned for December, with more than 500 pre-orders already registered on the company's website.
Updated on Nov. 11, 2026 in Artificial Intelligence
Live Poll
Would you use a voice AI wearable in noisy public places like cafes?
Voice AI users in San Francisco's noisy streets and cafés may soon have a wearable option built for them, with a local startup preparing its first public beta. Antimattr, founded in 2025 and based in San Francisco, is developing hardware to make voice AI work in loud settings.
Why it matters
Anyone who has tried to dictate a message on a crowded Muni platform or in a Mission café knows how unreliable voice assistants get when the background is loud. Antimattr's bet is that better hardware, not just better software, can make talking to AI practical in everyday city life, and it is still an early product whose real-world performance has yet to be proven.
Antimattr reports more than 500 customers have registered pre-orders through its website, alongside a Rs 25 lakh founder contribution and a Rs 20 lakh equity-free grant from WTFund. The company has not disclosed pricing, and the performance of its noise-isolation claims has not been independently verified.
The players
Antimattr
A San Francisco startup developing wearable hardware for voice AI in noisy environments.
Y Combinator
The startup accelerator that selected Antimattr for its programme and is backing the company.
WTFund
Nikhil Kamath's fund, which gave Antimattr a Rs 20 lakh equity-free grant.
The details
Project Mnemosyne pairs a smart ring with specialised earphones that use proprietary voice isolation technology to filter surrounding noise. The devices record only when a user manually activates them, and raw audio is not used to train AI models, according to the company. Antimattr is preparing to open a public beta programme in December and is working toward SOC 2 and HIPAA compliance.
Timeline
2025: Antimattr is founded by Sridipto Ghosh and Sirsha Chakraborty.
2026: The company is selected for Y Combinator's accelerator programme.
December: Antimattr plans to open its public beta programme.
Across the Bay
Antimattr is the first wearable technology company of Indian origin to receive Y Combinator's backing, placing a Bay Area hardware bet inside the accelerator that has launched many of the region's best-known startups. Its wearable approach to voice AI extends a software-first wave of assistants into a form factor that has yet to prove itself at scale.
If you have been waiting to try a wearable that talks to AI in a loud Mission café or on a busy street, a December public beta is the next date to watch. Pre-order holders and waitlist sign-ups should expect the rollout to start with early users before any broad release.
The takeaway
Watch whether Antimattr's noise-isolation claims hold up once the December public beta opens to users outside the company. Until independent tests or beta reviews appear, treat the 500-plus pre-orders as early interest rather than proof that the hardware works in practice.
What happens next
Antimattr plans to open its public beta programme in December, which will be the first chance for outside users to test its voice isolation claims.
Further reading
For more on where AI tools are heading, explore our Artificial Intelligence coverage.
Source note: This article includes information reported by NDTV Profit.
Live Poll
Would you use a voice AI wearable in noisy public places like cafes?
