Gopawe builds voice AI for African languages and accents, one country at a time. We start in Liberia with Koloqua, Kpelle and Liberian English, then Sierra Leone with Krio.
African languages and creoles are expressive, regional and often written to match their sound. We build spelling guides and reviewed examples with native speakers, so the models learn consistent patterns while respecting everyday speech.
Each country follows the same steps: choose the languages, build the text base, record speakers with consent, train, test with native speakers, then launch. Liberia is first; Sierra Leone is next.
Translates between English, Koloqua and Kpelle in every direction.
Reads English, Koloqua and Kpelle aloud with Liberian accents.
Listens to Liberian speakers and turns speech into text.
Recordings and corrections from speakers shape every model. The recording workflow includes consent steps, and voice material is only used for the purposes people agreed to.
These are early releases. Some abilities may be uneven across speakers, topics and regional styles. Release notes describe what changed and where each model is still weak.
