Reclaim your voice.
A wearable that reads slurred speech from your throat and says it clearly, in your own voice.
Millions of people, repeating themselves.
- people worldwide struggle to be understood when they speak
Voice assistants, captions and speech apps are trained on typical voices, but on slurred speech, they fail.
Built for people who still have a voice.
- People living with ALS
- People recovering from a stroke
- People living with Parkinson’s
- People who stutter
- Anyone whose speech is hard to understand
How Vocera works.
Throat vibrations
Sensors on your neck feel your voice box vibrate, catching sounds a mic misses.
Microphone
Picks up your words, like any mic.
AI trained on slurred speech
Learns from people with dysarthria, then keeps learning your voice.
Clear speech, in your voice
A speaker says your words clearly, so anyone can understand you.
Trained to understand slurred speech.
We shrank a top speech AI 10× to fit on a wearable, then trained it on speech from people with dysarthria.
- of words understood, after 3 hours of training on slurred speech
- faster training, from rewriting the larger model’s code
- to train and test everything, on one NVIDIA GB10 computer
It learns the patterns of impaired speech.
- ALS: “Can you help me sit up a little?”
- Stroke: “My daughter is coming over at five.”
- Parkinson’s: “Please leave the porch light on.”
- Whispering: “I’ll call you back after the meeting.”
- Noisy room: “Two coffees, one with oat milk.”
Gets more accurate every time you use it.
Hardware.
Scroll to build it.
Two 3D-printed parts
A pod for the throat, and a hub at the back of the neck to carry the weight.
Battery and brains in the hub
A thumb-sized ESP32 chip with WiFi, plus a battery that lasts 4+ hours.
Three sensors in the pod
Two piezo discs feel your voice box vibrate, and a mic hears the room.
A speaker that talks for you
A 13 mm speaker sits behind a grille on the front of the pod, facing out so whoever you’re talking to hears you.
Sealed and strapped
Lids close, a strap goes on, and one cable carries the signals to the hub.
- sensors, recorded in perfect sync
- updates a second, sent to the AI
- of streaming on one charge
It works where mics fail.
A mic only hears sound in the air, so noise drowns you out. Vocera also feels your voice as vibrations in your throat, which background noise barely touches. That makes it useful far beyond slurred speech.
Other potential applications
- Military and policeClear radio over engines and gunfire, or a whisper on covert missions.
- Firefighters and rescueTalk clearly through a breathing mask.
- Construction and factoriesBe heard over drills, engines and machines.
- Pilots and driversBe heard in cockpits, tanks, race cars and on motorcycles.
- HospitalsSpeak clearly through N95 masks and protective gear.
- Concerts and eventsStage crews can talk over the music.
- Private voice commandsWhisper to your phone or AI assistant on a train or in an office.
- Calls anywhereWind, traffic and crowds don’t get through.
- Gaming and streamingClean voice chat with no background noise.
- Speech therapyTrack speech progress at home between sessions.
Why Vocera?
| Product | Speech in | Portable | Non-invasive | Cost |
|---|---|---|---|---|
| AAC tablets | $3–8k | |||
| Voiceitt | $500/yr | |||
| Neuralink | Trial only | |||
| Vocera | ~$50 |
Built by
- Nikhil Bhimireddy
- Arav Maheria