What the Masakhane Playbook project has shipped, what's in flight right now, and what's planned for the coming quarters. Updated as work moves between columns.
Last updated: 2026-04-30
✅ Shipped (Q1 · 2026)
Foundations
- Foundational chapter scaffolding across the dataset lifecycle (10 sections, ~22 chapters)
- Open infrastructure: Docusaurus 3.10 site, downloadable PDF, RSS / Atom blog feeds, installable PWA for offline reading
- Internationalisation scaffolding for 6 languages (English + Hausa, Amharic, Swahili, French, Portuguese)
- Citation system: per-page "Cite this page" link, BibTeX, APA, MLA, Chicago, machine-readable
CITATION.cff - Community channels: Discord, GitHub Discussions, AfricaNLP Newsletter signup page
- Math (KaTeX), Mermaid-ready diagrams, comments via Giscus, Cloudflare Web Analytics
Community
- Case-study workshop on low-resource annotation — annotators, linguists, researchers, tool developers, and legal experts surfaced challenges and mitigation strategies for African data work
🟡 In flight (Q2 · 2026)
- Call for Chapter Development Proposals — open until 30 June 2026. Domain experts invited to lead chapters across text, speech, and vision/multimodal annotation. USD $1,000 honorarium per accepted chapter. Read more →
- Native-speaker review and translation of Hausa, Amharic, and Swahili chapter content
- Pilot deployments of the MasakhaneTool annotation platform at Bayero University and Bahir Dar University ICT4D for testing and validation
⏳ Planned (Q3 · 2026)
- First playbook release with 5 translated languages — culturally contextualised, community-reviewed first version published on Docusaurus, with downloadable PDFs for offline use
- Algolia DocSearch live — full-text search across all chapters and locales, free for OSS docs
- Tag pages and authors page polish for the blog
🌍 Planned (August 2026)
- Workshop at Deep Learning Indaba 2026 — 90-minute interactive showcase of the AfricaNLP Playbook and MasakhaneTool: live demos, hands-on annotation tasks, community dialogue. Pan-Atlantic University, Lagos, Nigeria. Workshop details →
🔭 Longer-term (2027+)
- Speech and multimodal annotation tooling (audio, speech-text alignment, image+text, video captioning)
- Benchmark integration with Lanfrica for dataset discoverability
- Versioned releases with Zenodo DOIs for stable academic citation
If you'd like to contribute to any of these workstreams, see the How to contribute section on the About page, or reach us on Discord.