antimattr is building voice AI wearables that cut through the noise


Founded in 2025 by Sridipto Ghosh and Sirsha Chakraborty, San Francisco-based deeptech startup antimattr is developing voice-first wearable technology that helps people interact with AI agents clearly, even in noisy environments. The startup was backed by Y Combinator and received a ₹20 lakh equity-free grant from Nikhil Kamath’s WTFund. Its early development was supported by Scaler Innovation Lab, which provided access to prototyping equipment.

Its first product, Project Mnemosyne, combines a ring and earphones. The earphones use proprietary voice-isolation technology to reduce background noise during calls and voice dictation, while the ring captures thoughts, controls AI agents and summarises conversations. Users can activate recording deliberately, and the company says it does not use raw audio to train its models.

The founders are also planning two more projects, Soteria and Athena, as part of a broader vision to create private, locally operated AI systems and an ecosystem of devices for natural human-machine interaction. The company is pursuing SOC 2 and HIPAA compliance to strengthen data security and privacy.

Antimattr says more than 500 customers have placed pre-orders through its website, with the public beta of Mnemosyne expected by December 2026. The founders initially invested around ₹25 lakh and believe their focus on making voice AI practical in noisy, real-world environments gives them an advantage over competitors such as Bee.