In Kenya, a young startup is taking on one of the most persistent challenges in communication: bridging the gap between spoken and sign languages. Founded in 2023 by 24-year-old entrepreneur Elly Savatia, Signvrse has created Terp 360, an AI-powered platform designed to translate text and speech into Kenyan Sign Language (KSL) using realistic 3D avatars.
From Idea to Innovation
According to an Africa News article, the idea emerged from Savatia’s firsthand awareness of the barriers faced by Deaf and hard-of-hearing communities in daily life. Whether navigating healthcare services, public offices, or the workplace, communication often depends on the availability of a human interpreter—a resource that can be limited or costly. By digitizing the interpretation process, Signvrse aims to make sign language translation more accessible, faster, and available on demand.
At the core of Terp 360 is motion-capture technology that records the movements, gestures, and facial expressions of fluent sign language interpreters. These recordings are then used to train AI models that animate avatars capable of delivering accurate and natural-looking KSL translations. On YouTube, Savatia describes the platform as a “Google Translate for sign language,” emphasizing its potential to work across a variety of everyday situations.
Early Adoption and Recognition
The platform is currently in a web-based beta phase, with around 2,000 active users, according to the Africa News article. For now, it translates from English into KSL, but expanding into other source languages — particularly Swahili — is a high priority. According to early users, this addition would significantly increase its usefulness in Kenyan contexts, where Swahili is a dominant language alongside English.
The initiative has already attracted national and international recognition. Signvrse received the Kenya Presidential Innovation Award and was selected for the Google.org Accelerator: Generative AI, a program that provides technical mentorship, cloud computing credits, and financial support. Participation in the accelerator is expected to help the team scale its technology, improve translation accuracy, and broaden its sign language coverage beyond KSL.
Looking Ahead
Not all experts are convinced the technology is ready for full-scale deployment. Some sign language trainers note that AI translations can still produce errors, and that current functionality benefits hearing users more than Deaf users. However, many agree that the potential is considerable, particularly as AI models improve and vocabulary databases expand.
According to the Africa News article, Signvrse plans to launch a mobile app by December 2025, introducing a subscription model for both individual and enterprise use. If successful, the project could not only reshape sign language accessibility in Kenya but also serve as a model for similar initiatives across Africa and beyond.
