YOUR WORK COULD MATTER HEREA little about
A little about
the possibility.
Most medical AI was built for English. We're fixing that. You'll research and develop NLP models for African languages — Swahili, Amharic, Yoruba, Hausa, Zulu, Kikuyu, and more — applying them to clinical conversation and patient communication contexts.
What you’ll work on
- Collect and curate African language medical text corpora
- Fine-tune multilingual models (mBERT, XLM-R, AfriBERTa) on African clinical text
- Build evaluation benchmarks for African language medical NLP
- Develop culturally-aware conversation flows for patient communication
- Research code-switching and dialect variation in clinical contexts
- Collaborate on African language AI with regional NLP communities
What you bring
- Strong background in NLP and computational linguistics
- Proficiency in Python and HuggingFace ecosystem
- Native or near-native fluency in at least one African language
- Understanding of multilingual NLP challenges
- Familiarity with language model evaluation metrics (BLEU, BERTScore)
A little extra, if you have it
- Contributions to Masakhane or similar African NLP initiatives
- Experience with speech recognition (ASR) for African languages
- Knowledge of low-resource NLP techniques
- Academic research background in linguistics or computational linguistics
What the role offers
- Remote-friendly work arrangement
- Co-authorship on NLP research papers
- Access to unique African language datasets
- Contribute to open-source African AI
Tools & skills
PythonHuggingFacemBERTXLM-RAfrican NLPMasakhaneMultilingual AI
Have a question before you apply?
Talk to the careers team