Regional Africa: Afri-Aya V2 Kickoff

​Afri-Aya V1 became the largest human-annotated image collection across 14 African languages , Gold Award at Expedition Aya 2025, and 162 HuggingFace downloads last month alone. Now we build V2.

In this session, community co-lead Bakunga Bronson, joined by members of the original Afri-Aya V1 team, covers what V1 taught us, the V2 roadmap , more languages, stronger quality control, native-speaker-led data collection from Kigali to Kano and the concrete ways members can take ownership: per-language data work, the research paper heading to publication, annotation quality, and community operations.

Whether you've been here for years or joined last week: if you've been waiting for a real way to contribute to African-language AI, this session is the door. All regions, all levels, all welcome. 🌍