Speech and language tooling
Translating Sauti Project research into production-ready tools - speech recognition for African languages packaged for use in apps, accessibility tooling, and field research.
TRI AI Innovation
Turning research into things that ship.
TRI AI Innovation is the youngest of our three divisions. Its job is to take the work emerging from Teaching and Research and translate it into products, tools, and spinouts that serve African contexts directly.
Currently in development · First launches: late 2026
01 - Status
In development.
We're prototyping our first launches across three areas - speech and language tooling, computer vision applied to African contexts, and education infrastructure. The first product release is targeted for late 2026; until then, this page is the public side of work that's happening internally.
If you're a researcher, engineer, or organisation interested in collaborating on an Innovation initiative - or you'd like to hear when the first products launch - get in touch.
02 - Focus areas
What we're building.
Three areas where TRI AI's research and teaching naturally extend into shippable work.
Vision systems for African contexts
Building on ChowNet, applied vision systems for domains where mainstream models fail - food, agriculture, marketplace inventory, healthcare imaging.
Education infrastructure
Internal tooling that scales TRI AI Saturdays, watch parties, and the colloquium series - and which we plan to release publicly so other African AI communities can run their own programmes.
Newsletter
The Encoder.
Monthly programmes, research, and opportunities updates from The Encoder — a TRI AI Initiative