The people building Somali AI
Every dataset, standard, bootcamp, and open-source project behind Somast was shipped by its contributors: 275 of them.
Counted across 13 public GitHub repositories. Snapshot taken 2026-09-02.
Recognising the work that moved a project forward
Shipped the maanka2 Somali Web Corpus downloader into Somali NLP Corpus on 1 September, a new source feeding the text pipeline the rest of the corpus work builds on.
Maintainers of the core repositories
Commit access, review duty, and release ownership across the corpora, the bootcamps, and the standards repos.
Ranked by commits, all repos
Commit counts are one signal among many. Review load and annotation hours do not show up here.
Everyone who has shipped something
Code, data, docs, and bootcamp work. Every merged contribution counts the same on this page.
Where the work happens
How to get on this page
We build with the community. Annotation, transcription, and translation count as much as code.
Every repo keeps an open queue. Start with one sized for a first pull request, no prior lab context needed.
Draft pull requests get reviewed quickly, and questions in Somali on the issue thread are welcome.
Merged work adds you to this page, to the repo contributor list, and to the dataset acknowledgements.
Build Somali AI with Somast
Use our published datasets and models, contribute to the research, or partner with us.
Researchers
Use our datasets and benchmarks, co-author studies, and push low-resource NLP forward.
Developers
Build on published models and tokenizers, file issues, and ship Somali-language features.
Students
Join fellowships and bootcamps to learn ML and contribute to real research.
Contributors
Help collect, transcribe, and verify data. Every contribution improves the models.
Need software built, or want to partner with us? Services · Partners