Community-Centered Voice AI Workshop for the Dagbani Language

The Community-Centered Voice AI Workshop was successfully held on 31st January and 1st February 2026, bringing together Dagbani language volunteers to contribute to the development of inclusive voice technologies. The workshop focused on building participants’ practical skills in sentence annotation and sentence validation on the Mozilla Common Voice platform.

Over the two-day training, participants were introduced to the goals of community-driven language data collection and the importance of high-quality, locally generated datasets in the development of speech recognition technologies. Facilitators guided volunteers through hands-on sessions on how to accurately annotate Dagbani sentences, validate existing entries, and follow quality and consistency standards required for open voice datasets.

The workshop emphasized the role of community participation in ensuring that underrepresented languages such as Dagbani are adequately represented in emerging artificial intelligence systems. By equipping local speakers with the necessary technical skills, the initiative contributes to the preservation, digitization, and technological relevance of the Dagbani language.

This workshop forms part of a broader effort to promote linguistic diversity in artificial intelligence and to empower local communities to take ownership of their languages in digital spaces. The training also fostered collaboration among volunteers, encouraging knowledge sharing and sustained engagement with the Mozilla Common Voice platform beyond the workshop period.

The project is funded by An AI of Our Own and implemented in partnership with Mozilla Common Voice and the Dagbani Wikimedians User Group. Through this collaboration, the initiative continues to strengthen community capacity and advance the inclusion of African languages in open, community-driven AI projects.