I’ve spent months testing dictation apps, and most of them feel like a compromise—great in a quiet home office, a social liability on the train. Mina Fahmi gets that. The Sandbar CEO built the Stream, a smart ring shipping later this year, because he was tired of digging out his phone mid-walk every time an idea struck. It’s a refreshingly introverted gadget: a microphone, a battery, and Bluetooth, all focused on capturing your voice, not recording everyone else’s meeting.
In a New York demo, Fahmi did something I haven’t seen work this smoothly before. He flashed new firmware onto the ring, connected it to his Mac and iPhone, and dictated a couple of sentences into Apple Notes by simply holding the ring to his mouth and pressing a touchpad. The transcription appeared, accurately, on both devices. He wouldn’t explain the plumbing, but watching him switch seamlessly between phone and laptop without touching a keyboard felt like a genuine leap beyond the usual AI gadget demos that just end up as a bad smartphone app.
Fahmi is positioning the Stream for three specific jobs: firing up an AI assistant, jotting notes, and dictating text directly into any text field. The current companion app is a stopgap. The real promise—and the hard part—is making this work inside your existing tools. The ring even has a whisper mode that picked up Fahmi’s voice when I couldn’t hear a word he said. At least, he says it did.
Plenty of companies are chasing this space, from Pebble founder Eric Migicovsky’s Index 01 to the meeting-focused Vocci. The tech isn’t exotic. It’s a mic. What’s clever is the form factor. You don’t have to unlock a phone or open an app. You bring your hand to your mouth, a gesture we all understand, and you talk. It’s a far cry from shouting at your headphones in public, and it might finally be the AI hardware that doesn’t make you look like you’re performing for an audience.
