Live Meeting

Everyone in
the same room.

SignLine is real-time, two-way video calling with built-in sign-language translation. Deaf and hearing participants communicate seamlessly in the browser — no interpreter on standby, no software to install.

At a glance

Participants per room

P2P

WebRTC mesh — no media server

2-way

Sign ⇄ speech translation

100%

In-browser — no install

How it works

Create a meeting, share the five-letter code, and your peers join straight from a browser. SignLine connects everyone over a peer-to-peer WebRTC mesh, so video and audio flow directly between participants. A Deaf participant signs; their words are recognized on-device and delivered to hearing peers as live captions and natural speech. A hearing participant speaks; their words are composed into a sentence and rendered as a signing avatar for the Deaf participant.

Live captions accumulate into an editable history, sentences are composed with the Gemini LLM for grammatical fluency, and an optional background blur keeps your surroundings private.

Use cases

Team standups

Mixed Deaf and hearing teams meet without scheduling an interpreter for every quick sync.

Telehealth

Patients and clinicians converse directly, with captions attached to the consultation record.

Education

Remote tutoring and group study where every participant follows along in their own modality.

Family calls

Stay close with Deaf and hearing relatives in one call — everyone understood, no middle-man.

Technical architecture

SignLine uses a full-mesh WebRTC topology: each participant holds a direct peer connection to every other participant, with Socket.IO handling only signaling (offer/answer/ICE). There is no media server, so video never leaves the peer-to-peer path. The newcomer deterministically offers to everyone already present, which avoids glare, and connections renegotiate when a participant toggles their camera mid-call.

Recognition reuses the same on-device MediaPipe + TFLite pipeline as the rest of Together; sentence composition routes through the Gemini LLM, and optional background blur runs MediaPipe Selfie Segmentation locally on the outgoing track.

Start a SignLine call

Open the dashboard, choose Live Meeting, and share your room code. No account required for a demo.