Sharing Language AI Expertise at KNUST

African Languages Lab delivered a lecture at Kwame Nkrumah University of Science and Technology (KNUST), hosted by the university's Responsible AI Lab (RAIL). The talk, given by our founder Sheriff Issaka, was titled Building AI Solutions That Work for Africa.

Africa's advantage

The lecture opened with a simple point: Africa's young and growing population is an advantage for AI innovation. The people who will build, use and shape AI on the continent are students today, and the problems worth solving are the ones they see every day, starting with language.

Small models, big impact

We discussed small language models built for languages with limited data. Focused models can deliver strong results for specific languages and tasks, while staying affordable to run on the infrastructure many African organisations actually have.

Three practical priorities

• Better recognition of underrepresented languages, so AI can hear and read the languages people actually use.

• Inclusive multilingual systems that handle dialects, accents and code-switching.

• Protecting cultural heritage through digital tools, so spoken knowledge is preserved, not lost.

Mansa Translate in action

The lecture included a look at Mansa Translate and how it is used across education, media and business, including how its tone settings (Precise, Natural, Expressive and Formal) help preserve meaning for different audiences.

Why we do this

African languages and knowledge systems belong at the centre of global AI conversations. Working with universities like KNUST is how we grow the next generation of researchers building for the continent. Our research collaborations have already mentored 15 early-career researchers, and students who want to get involved can explore our fellowships and internship program on the Opportunities page.