Research
I research natural language processing for underserved languages. Right now that means Somali: building the datasets, tools and models a language needs before AI can serve the people who speak it.
NLP for underserved languages
Natural language processing for the languages the big models leave behind. The gaps aren't about how many people speak a language; they're about who built the tools.
Somali, specifically
Over 20 million people speak Somali, yet it has almost no usable AI: no solid speech models, no quality datasets. That's the gap I'm working in.
Tools and models
Building the pieces a language needs before AI can serve it — datasets, speech and language models, and the tooling to keep them alive.
Working on the same languages, or want to help? I'd like to hear from you — abdulebaroo@gmail.com.