Join Dr Shamsuddeen Muhammad as he explores how AI can serve African languages and the communities that speak them.

Field recording, speaker recruitment, and quality control — the practical side of building ASR datasets when there is no pre-existing corpus to start from. With a guest researcher from EthioNLP.

A walk through the last five years of Hausa NLP — from early word embeddings to current LLM evaluation benchmarks. What works, what is still missing, and what newcomers should read first.

Why treating annotators as collaborators rather than crowd workers changes the quality of African-language datasets. Conversation with a Yoruba Language Lead from the AfriFinder pilot.

In the opening episode, Dr Shamsuddeen Muhammad lays out the data, modelling, and community gaps that hold African-language AI back — and what realistic next steps look like.