From camera motion to shared language.

Built for Deaf and signing communities, interpreters, platforms, and public services — as software infrastructure, not a learning catalogue.

Mission

Make every sign language usable in software.

Digest motion into text and text back into motion, so signed communication works wherever people already work — browsers, mobile apps, interpreting workflows, and client platforms — with privacy as a default.

Bidirectional core

The platform turns signs into software.

One engine for any sign language. Not a study site. Recognition and expression grow as models and language coverage mature.

Motion → text

Signed meaning, written out

Motion sequences are recognized and mapped to written or spoken meaning for that sign language.

Text → motion

Written meaning, signed back

Text maps into signing motion for avatars, accessibility flows, and client delivery.

Full vision

Layers that stay replaceable.

01 Perceive On-device UniversalMotionFrame. Video stays on the device.

02 Digest Engine maps motion ↔ text for every supported sign language.

03 Deliver Text, speech, APIs, embeds, SDKs into Neo and other clients.

04 Express Mapped text drives articulator motion on an avatar.

Today

What ships on this origin.

Perception client, motion-frame contract, live console, embed surface, and articulator avatar. Bidirectional motion ↔ language for any sign language is the core product layer ahead.

Subscribers Console On-device tracking and motion JSON — sign in required. Subscribers Avatar Articulator commands for hands, face, and torso — sign in required. Subscribers Client integration API keys, origins, and embed snippet in the account panel.

NEOLANG

Built to power NEOLANG.

SignToTalk is the sign perception and motion ↔ language layer behind NEOLANG — on-device motion tracking, signed motion to text and back, and embeds that keep the camera on the user’s device. Explore the platform at neolang.de.