Zimbabwean startup Bhala has launched a mobile virtual keyboard that uses machine learning and state-of-the-art deep learning models to improve the experience of writing in African languages.
Rolled out in January, Bhala currently supports word auto-complete, spellcheck, and word prediction in ChiShona, IsiNdebele, IsiXhosa, IsiZulu, KiSwahili, Sepedi, Sesotho and SiSwati, with more coming soon.
Director of operations Njabulo Sandawana claims it has more accurate autocorrect and predictions than any comparable app on the market, including those Google’s Gboard and Microsoft’s SwiftKey.
“Too often Africans have to tolerate technology not designed for them or their needs. Bhala is different. It’s built by Africans, for Africans, and for African needs, while respecting privacy and ethical considerations through research-based methods,” he said.
“We are committed to providing technology that enables users to communicate effectively, efficiently, and improves their user mobile device experience.”
The self-funded platform currently has more than 10,000 downloads, and has already become one of the top free apps in Zimbabwe and eSwatini.
“Since the release of our beta we have been focused on the Zimbabwean market, but we are extending to include more countries.The goal is to bring Bhala’s language technology to the entire continent,” said Sandawana.
“Our goal is to grow as fast as possible. We are focused on onboarding as many users as possible and on constantly improving our technology for the benefit of our users.”