Building software for multilingual video localisation.
Hatek-Lingua AI is a development-stage AI product focused on translating spoken video, generating target-language speech and synchronising visible speech to the translated audio.
Make useful video easier to understand across languages.
A lecture, interview, training video or documentary should not need a completely new production process for every audience that speaks another language.
The mission is to reduce the technical and production work required to create high-quality multilingual versions of existing video, while paying particular attention to African languages that are often less well served by current tools.
A software workflow that combines language and video AI.
The product is intended to let a user provide a source video, choose a target language, generate a translated audiovisual version, review the result and export the approved output.
Source video
The original content provides the speech, timing, speakers and visual performance.
Translation
The spoken message is converted into the selected target language.
Generated speech
Target-language audio is produced with appropriate pronunciation and timing.
Video synchronisation
The visible speaker is adjusted to follow the new speech.
Review
The final output is checked before use or distribution.
Language coverage is uneven, especially when speech and video quality are considered together.
Some languages have abundant speech and translation resources; others do not.
Hatek-Lingua AI is intended to invest specifically in the data, model adaptation, pronunciation work and native-speaker evaluation required to improve African-language audiovisual localisation rather than treating those languages as an afterthought.
Shola Fatusin and Kaori Fatusin.
Shola Fatusin is the Founder of Hatek-Lingua AI. Kaori Fatusin is the Co-founder.
The company is building the technical, language and operational capacity required to move the product from research and development into validated pilots and commercial use.
The product requires expertise across several disciplines.
Relevant areas include machine learning, speech, translation, computer vision, data engineering, language research, product engineering and quality evaluation.
Researchers, language specialists and technical collaborators who can contribute to those areas can use the contact page to introduce themselves.