We started Gestura because the gap between hearing and Deaf conversations is closed every day by people, not machines, and those people get tired. Interpreters cost money. Family members get pulled into doctor visits. Friends learn three signs and stop. The fluent middle is missing, and phones are everywhere.
So we asked a small question. Can a modern phone, looking through its own camera, do useful sign translation in real time, on device, without sending video anywhere?
After two years of stubborn iteration, the answer is yes. Not in every scene, not in every light, not perfectly. But in everyday rooms, with everyday people, at conversation speed, with the model running entirely on the phone, the answer is yes.
What ships first
The first build of Gestura does two things.
- Translate Croatian Sign Language to text in real time, from your camera, on device.
- Teach the same language back to you, gamified, in short daily sessions.
One app, two directions of the same bridge.
What is next
Slovenian and Austrian sign languages are next on the roadmap. The architecture is language-agnostic; the bottleneck is data, collected together with the Deaf community, never scraped.
We will be honest about what works and what does not. Expect more posts here, short, often, and specific.
If you want the beta, join the waitlist. Twenty-eight hundred people are already on it.