Explore the research behind our models. We are committed to transparency and community-driven development in West African AI.
We present a novel approach to building speech-language models for West African languages, focusing on data efficiency and cultural relevance.
Firtanam is an open-source speech recognition system optimized for the Fulani language family, supporting multiple dialectal variations.
An exploration of how scaling laws differ when training language models on predominantly oral languages with limited digital resources.
Technical architecture for deploying voice AI in low-connectivity environments, with a focus on edge computing and model compression.