Why I’m Building ViaSign

I am a Deaf and hard-of-hearing Singaporean developer. I started building ViaSign because one question kept returning to me:

What would accessibility technology look like if signed language were part of the foundation, rather than something added near the end?

Captions, transcripts, and speech-to-text matter. I use and value tools like these. But access is not one format, and written language is not always the same thing as language access.

For people who communicate through signed languages, access can also mean being able to receive information in a language that uses the hands, face, body, timing, and space together. That is a different design problem from placing English words on a screen.

The gap is about language

Signed languages are languages, not spoken sentences with each word replaced by a gesture. Meaning can depend on sentence structure, movement, direction, facial expression, body position, spatial reference, and context.

That means a system can produce something that looks like signing while still carrying the wrong structure or meaning. Smooth animation cannot correct a weak language foundation. It may only make the mistake look more convincing.

ViaSign begins with Singapore Sign Language because that is the language context closest to this work. It does not claim to define SgSL, replace interpreters or teachers, or act as an official language authority.

Structure before spectacle

The most visible version of sign-language technology is often an avatar. I understand the appeal: movement makes the idea feel real immediately.

But an avatar should not move before the system can explain what it thinks the sentence means, which structural decisions it made, and where it is uncertain. That is why the wider ViaSign research direction starts further upstream, with grammar research and inspectable intermediate plans before motion.

The current public parser is a narrower safety-first boundary. It exposes a few closed natural-English surface categories and visible review states; it does not implement SgSL grammar, translation, sign plans, or motion. Earlier private parser research explored possible features such as sentence function, time, negation, and reference, but those notes are not shipped capability or final SgSL. Uncertain decisions must remain visible as requiring review rather than being hidden behind confident motion.

The deeper engineering direction is documented in the parser notes and the ViaSign roadmap.

Review belongs inside the system

I can bring my own Deaf and hard-of-hearing lived experience to this work, but I do not represent every Deaf person, signer, teacher, learner, interpreter, or SgSL user.

That boundary matters. Building from lived experience is not the same as having community authority. The project needs paths for people with relevant language and community knowledge to inspect, question, correct, and reshape what the system produces.

ViaSign does not yet have all of those relationships or processes in place. The community review process describes a direction to build toward, not a claim that formal community validation has already happened.

What Viathorne Labs means today

Viathorne Labs is the public name I use for this work. It is an early-stage research and technology project, not yet an incorporated company.

I still want to build it with the seriousness I would expect from a future company: clear boundaries, inspectable decisions, honest prototype labels, accessible communication, and evidence before claims.

Becoming a company would be a milestone, not something a website can declare into existence. The work has to earn that next step.

The future I am working toward

I hope ViaSign can grow into a useful bridge: a system that helps people explore how meaning might move from English input toward reviewable SgSL structure and, only when the foundations are ready, toward careful motion research.

That will take time. It will involve mistakes, technical limits, better sources, and feedback that changes the direction of the project. Some ideas may not work the way I first imagine them.

What matters is that the uncertainty stays visible and that progress is not measured only by how impressive the output looks.

I am building ViaSign because signed-language access deserves sustained, careful work—and because I believe technology can support communication without pretending to replace the people and knowledge that make language alive.

You can read the current public boundary in Why Sign Language AI or follow the evolving project from the ViaSign overview.