The story
YourStory covered Vitra.ai in January 2024 under a framing the company has used since its founding: most of the world's best content exists in a language most of the world does not read. Video and podcasts have made that gap worse, not better, because audio and moving images are far more expensive to localise than text.
Vitra.ai's answer is to treat translation as a pipeline problem rather than a linguistic one. Transcription, translation, voice synthesis, lip synchronisation and subtitle burn-in are separate models chained together, so a creator uploads a file and receives a finished, watchable asset rather than a transcript they then have to produce against.
Why accessibility is the right frame
Framing this as accessibility rather than marketing reach changes what "good" means. A dubbed video that is fluent but wrong is worse than no translation at all, which is why the engine is context aware and why enterprise customers in regulated industries run output through quality control before publication.
The same pipeline is what powers video dubbing for enterprise customers today.
Source
- Your Story — Breaking the language barrier: Using AI to make videos, podcasts accessible