What happened
Sanzhar Yerezhepov, a 15‑year‑old student at the Daryn gifted‑school in Kokshetau, Kazakhstan, has created Tusin, an artificial‑intelligence‑powered translator that turns spoken or written content into lifelike sign‑language videos. The system generates photorealistic avatars that reproduce hand movements, facial expressions, and emotions, aiming to improve accessibility for deaf and hard‑of‑hearing Kazakhstani citizens. Tusin is currently being tested with the Kazakhstan Society of the Deaf and has attracted interest from five organizations. The project advanced to the semifinals of an IT competition at Astana Hub, the country’s startup hub, and Yerezhepov reports plans to expand the service across the Commonwealth of Independent States (CIS).
Yerezhepov conceived Tusin after observing deaf students at a nearby boarding school struggle to access online information. Working with the Kazakhstan Society of the Deaf, he built a prototype that converts audio or text into video clips featuring a photorealistic signer. The system incorporates AI techniques to interpret context, allowing the same sign to convey different meanings depending on surrounding words.
The platform currently supports Kazakh and Russian spoken input, translating both into Russian Sign Language. Yerezhepov says the AI can also recognize English speech and render it in Russian Sign Language, with plans to add other sign languages such as British and American Sign Language if the product reaches international markets.
Tusin has attracted interest from five organizations and reached the semifinals of an Astana Hub competition, indicating early market validation. Yerezhepov’s team includes his computer‑science teacher and mentors from local incubators like Startup Orda 2.0 and the Social Impact Award Kazakhstan program.
Why it matters
Sign‑language translation differs from subtitles because it conveys grammar, facial expressions, and cultural nuances that many deaf users rely on. In Kazakhstan, Russian Sign Language is the primary sign language, and subtitles in Kazakh or Russian often fail to meet the needs of sign‑language‑native users. Tusin’s AI‑driven approach could fill a gap in the CIS market where comparable photorealistic solutions are absent, potentially influencing how governments, broadcasters, and private companies meet emerging accessibility requirements. The recent inclusion of sign language in Kazakhstan’s Social Code may spur regulatory pressure for broader adoption, making Tusin a timely tool for compliance and corporate social responsibility. Moreover, the project demonstrates that advanced AI applications can emerge from youth‑led initiatives in regions outside traditional tech hubs, highlighting the growing global diffusion of AI talent.
Accessibility: By providing video‑based sign‑language translation, Tusin addresses a critical accessibility gap that subtitles cannot fill, especially for users whose primary language is sign language.
Regulatory momentum: Kazakhstan’s recent legal recognition of sign language may soon translate into mandatory accessibility standards for businesses, creating a market pull for solutions like Tusin.
Regional impact: The CIS lacks comparable photorealistic sign‑language tools, so Tusin could become a de‑facto standard across multiple countries that share Russian Sign Language.
Youth‑driven innovation: The project showcases how young talent in emerging economies can develop sophisticated AI products, potentially encouraging more investment in local AI ecosystems.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').Which component of an AI application is the machine-learning model itself?
What to watch next
Key indicators to monitor include: (1) whether Tusin secures formal contracts with government agencies or media outlets required to provide sign‑language interpretation; (2) the pace of scaling beyond Kazakh and Russian to other CIS sign languages; (3) any regulatory changes that mandate sign‑language accessibility for private firms; and (4) potential partnerships with larger tech firms or investors that could accelerate deployment and improve the underlying AI models.
Contract announcements from Kazakhstani ministries, broadcasters, or transport operators that would require sign‑language video services.
Funding rounds or strategic partnerships with larger technology firms that could scale the AI model and production .
Legislative updates that expand the legal obligations for private sector accessibility, which would increase demand for Tusin.
Technical milestones such as the addition of new sign languages, improvements in AI context handling, or the release of a public API.