Nirva is developing an always-on pendant that it says listens for its wearer’s speech, then uses the information to offer personal insights. The startup has raised $33 million and plans to begin taking orders in November, with shipments to early users targeted by the end of the year.
A wearable designed for everyday use
Nirva announced a $25 million pre-Series A round, adding to the $8 million it raised earlier in the year from investors including Lightspeed and South Park Commons. Founded in 2025, the company was started by former Meta Reality Labs employees Wei Lyu and Jason Chen, alongside Hanying Hu, who founded a fashion and jewelry brand.
The first device is a pendant that can also be worn as a bracelet. Nirva says it weighs around eight grams and can last up to three days on a charge. The company describes the design as jewelry that can blend into daily wear while providing a continuous record of the wearer’s own speech.
The device is priced at $249 and is planned in silver, brushed silver, black, and gold. Nirva says it hopes to develop more devices over time, though the pendant is its first product.
How Nirva says its listening works
Continuous listening raises a basic privacy concern: people nearby may worry that their conversations are being captured. Nirva says it addresses this by asking users to record voice snippets during onboarding, then building a model to recognize the wearer’s speech and filter out conversations or audio from others around the device.
The company says it does not keep audio or transcripts by default after processing. Before discarding audio, it can track qualities such as tone, pitch, volume, and speed, which Nirva says help it derive emotion-related insights. The source does not describe the technical details of how the filtering or analysis works.
That distinction is central to Nirva’s pitch. CEO Wei Lyu told TechCrunch that the product is intended to focus on the wearer’s own speech rather than act as a general voice recorder or meeting note taker. It is designed to listen passively, rather than require the user to start a recording for each conversation.
Personal insights, reminders, and emotion
Nirva positions itself as a personal insights company rather than a productivity startup. Its app analyzes the wearer’s speech and presents patterns it associates with personality, people who often appear in the user’s life, places they have been, and other daily notes.
The app also provides reminders and action items. Lyu said the aim is to help people show up better for others, while supporting self-understanding, self-compassion, and self-awareness. The company’s example is a reminder prompted after a user forgot to thank her mother for something.
These features make the product’s interpretation of speech consequential. Nirva says it surfaces observations about emotions and communication style, as well as suggestions for improvement. The company describes those observations as insights inferred from everyday speech, rather than just a transcript or a list of tasks.
A crowded field with a distinct promise
Nirva is entering a growing category of AI wearables and ambient devices. The article points to Taya, which is also developing intelligent jewelry, as well as Friend and Bee, which are pursuing companion gadgets. The product also arrives amid other efforts to make AI features available through wearables and phones.
Nirva’s differentiating claim is that its pendant listens only to the person wearing it, while using that speech to build personal insights. Whether users are comfortable wearing an always-listening device will depend in part on whether they trust its filtering and data handling. For now, the company’s stated approach is to identify the wearer from onboarding voice samples, filter out other people’s speech, and discard audio and transcripts after processing.
Orders are expected to open in November, with early-user shipments aimed for the end of the year. That launch will put the company’s privacy claims and its promise of useful personal insights into practice for users.