DSFSI at Deep Learning Indaba 2025

We are excited to share that the Data Science for Social Impact (DSFSI) research group will be making multiple contributions at this year’s Deep Learning Indaba 2025. Our team will be showcasing research, leading conversations, and helping shape the future of African AI across key themes.

Posters

• Thapelo Sindane will present his poster “Injecting explicit cross-lingual embeddings into pre-trained multilingual models for code-switching detection”. His work explores how cross-lingual models can better capture the complexities of code-switching in African contexts.

• Nontokozo Manukuza will present her poster “Teaching AI to Read Between the Lines: Interpreting isiZulu idiomatic expressions through retrieval-augmented fine-tuning”. This research addresses one of the hardest challenges in African NLP: idiomatic language.

Workshops

• Thapelo is also part of the organising team for the Centring Data in African AI Workshop. The workshop will examine both technical and socio-technical aspects of African data, highlighting how ownership, context, and ethics shape AI for the continent.

• Within the same workshop, Prof. Vukosi Marivate will join the closing panel , sharing insights on building resilient African data ecosystems.

Panels & Keynotes

• Prof. Vukosi Marivate will also speak on the AI Policy in Africa Keynote Panel (Wednesday) and open the Voices of Africa 2: Building the infrastructure for African NLP Workshop (Friday).

And that’s not all — the DSFSI team is involved in much, much more at Indaba. Look out for Dr Idris Abdulmumin, Dinorego Mphogo, Takura Wekwete, Seani Rananga, Dawit Shibabaw and Fiskani Banda.

These are just a few highlights — the DSFSI team will be contributing in much, much more ways at Indaba. We’re proud to continue our mission of putting African languages, data, and innovation at the centre of AI.

🚀 Stay tuned for updates, and if you’re at Indaba, we’d love to connect!

👉 Learn more about the Indaba here: deeplearningindaba.com