- TwinMind passively captures speech to create AI-powered notes and to-dos.
- Ear-3 model supports 140+ languages and multiple speakers with real-time transcription.
- The app prioritizes privacy by storing text locally and deleting audio on the fly.
Former Google X scientists are turning science fiction into reality — or at least the kind that fits in your pocket. TwinMind, a startup co-founded by Daniel George, Sunny Tang, and Mahi Karim, offers an AI-powered app that passively listens to your conversations, meetings, and lectures to help you remember and organize information more efficiently. With $5.7 million in seed funding and both Android and iPhone apps now live, TwinMind is poised to reshape how we capture and use knowledge.
How TwinMind Works
Unlike conventional AI note-takers like Otter or Fireflies, TwinMind runs in the background all day (with user permission), building a personal knowledge graph from ambient speech. The app processes audio in real-time on-device, transcribes it, and can run continuously for 16–17 hours without draining battery life. It even supports offline usage and real-time translation in over 100 languages. Users can back up their data securely, though TwinMind does not store audio recordings permanently, prioritizing privacy.
TwinMind also offers a Chrome extension that gathers context from browser activity. Using vision AI, the extension can analyze emails, Slack messages, and Notion documents. In a practical example, the TwinMind team used the extension to sort over 850 internship applications, demonstrating the app’s potential for both professional and personal productivity.
Advanced AI Model and Features
TwinMind recently unveiled the Ear-3 AI model, a successor to Ear-2, which recognizes multiple speakers and supports over 140 languages. Ear-3 runs on the cloud but automatically switches to Ear-2 offline, ensuring uninterrupted service. The model is available via API for developers and enterprises, with a new Pro subscription offering extended context windows and faster support.
The startup currently serves over 30,000 users worldwide, with strong traction in the U.S., India, Brazil, and Europe. Users range from professionals and students to individuals leveraging the app for personal projects — even autobiographies.
From Google X to AI Startup
Daniel George’s experience at Google X, combined with his AI research background in astrophysics, has allowed TwinMind to launch rapidly from concept to product. Early investor Stephen Wolfram, along with Sequoia Capital and Streamlined Ventures, helped propel the company to a $60 million valuation post-seed.
Also Read: Google’s Gemini AI Rated High Risk for Kids, Warns Common Sense Media
TwinMind exemplifies the next wave of AI productivity tools — blending real-time audio capture, offline processing, and smart context generation — essentially giving users a “second brain” that is both practical and privacy-conscious.
Disclaimer: The information in this article is for general purposes only and does not constitute financial advice. The author’s views are personal and may not reflect the views of CoinBrief.io. Before making any investment decisions, you should always conduct your own research. Coin Brief is not responsible for any financial losses.
