Engagements

I'm the AI (strategy) lead of Wikitongues↗, a nonprofit working to sustain and document every language in the world. The initiative I run builds AI that speaks endangered languages authentically - and the first community-owned benchmark for how well frontier models actually speak them.

the benchmark is the lever

About 7,000 languages are spoken today - roughly half endangered, under 5% represented in AI. Off-the-shelf models answer confidently and get them wrong. A community-owned leaderboard - measuring how well ChatGPT, Gemini, and Claude actually speak a given language - is the lever that pushes the labs to do better.

the igala pilot

The first pilot is Igala (Yoruboid, ~2–3 million speakers, Kogi State, Nigeria), led in-community by Agnes and the Igala Wikimedians. Today's frontier models fail Igala badly, often confusing it with neighbouring languages.

learner, annotator, researcher

The platform has three roles. A learner chats and gets culturally-appropriate answers with a confidence score. An annotator - a fluent speaker - does pairwise A/B comparisons, rubric scoring, and free-text edits. A researcher reads leaderboards, rubric scores, and inter-annotator agreement. Behind it, a translator → reviewer → orchestrator pipeline scores confidence: high-confidence answers auto-pass, low ones route to human annotators, against an eight-part rubric spanning authenticity, cultural knowledge, idiom, register, lexicon, grammar, orthography, and dialect. The data stays community-owned and feeds a fine-tune / RAG / DPO flywheel.

funding & the ghana launch

$25,000 raised for the first three months, with collaboration and support from Google Research (Impact Lab), DAIR, and Georgia Tech, alongside academic advisors at NYU, JHU, and Sydney. The roadmap: a public launch and the first-ever Igala leaderboard at the Wikimedia Foundation conference in Ghana, October 4–5 2026.