Research
Research directions
What we are actively working on. Papers, datasets and model cards will be listed here as they are released.
01
Speech technology for Karakalpak
Text-to-speech and voice cloning are released as an API; speech recognition is in active research.
Active
02
Vision-language models for city safety
Scene and event understanding in camera feeds, beyond object detection.
In development
03
AI in public service delivery
Kiosks, notification systems and workflows that reach citizens directly.
In production
04
Karakalpak language data
Collecting, cleaning and licensing the speech and text data the language does not yet have.
Ongoing