Silenzio — Speech, without sound.
Vision Technology Principles Careers
The idea

Your voice starts before sound.

We capture the muscle activity behind articulation and decode the words already there.

How Silenzio works

From muscle activity to language.

A lightweight surface-EMG interface senses subtle facial and neck activity. A neural decoder maps those patterns into speech without requiring an audible voice.

01 / SENSE

Capture

Non-invasive electrodes record the electrical activity generated by the muscles involved in articulation.

02 / DECODE

Decode

Machine-learning models learn the temporal structure of silent articulation and map EMG patterns into linguistic representations.

03 / COMMUNICATE
OUT

Speak

Decoded language can appear as text, control a device, or be synthesized as speech — while the user remains silent.

A new interface for human expression.

01

Non-invasive first

Surface electromyography keeps the sensing layer outside the body while preserving the temporal information contained in muscle activity.

02

Language, not gestures

The goal is not just a small vocabulary of commands. We are building toward fluid silent communication that preserves the structure of natural speech.

03

Built around people

Comfort, privacy, calibration burden and day-to-day usability are treated as core engineering constraints

Silenzio

Make silence
speak.

We are building a silent-speech interface from surface EMG. If you work on neural interfaces, speech decoding, wearable sensing, or assistive communication, we would love to hear from you.