Access to reliable healthcare information remains one of the biggest challenges across many African communities, particularly in underserved and low-resource settings where language barriers limit access to accurate medical guidance. To address this gap, the Multilingual Health Question Answering in Low-Resource African Languages Challenge has officially launched on Zindi, inviting AI innovators, developers, researchers, machine learning engineers, and language technology enthusiasts to build multilingual AI systems capable of answering health-related questions in African languages.
The challenge is organized in collaboration with the Hub for Artificial Intelligence in Maternal, Sexual and Reproductive Health (HASH) and focuses specifically on maternal, sexual, and reproductive health information in African communities. Participants will work on creating AI models that can understand and generate accurate, fluent, and culturally appropriate responses in languages such as Akan, Kiswahili, Luganda, and Amharic.
The competition offers a total prize pool of $5,000 USD and is fully online, making it accessible to participants from across the globe.
Why This Challenge Matters
Despite the rapid advancement of artificial intelligence and natural language processing technologies, most large language models are still heavily trained on English and other high-resource languages. As a result, millions of African language speakers remain underserved by AI-powered healthcare systems.
This challenge seeks to change that reality by encouraging the development of multilingual health assistants capable of supporting African communities in their native languages. Such systems could eventually power:
- Rural health worker assistants
- Community healthcare support systems
- Maternal health education tools
- Adolescent sexual health platforms
- HIV and STI awareness systems
- Clinic support applications
- Patient education chatbots
By enabling people to ask sensitive health questions privately and receive responses in their own language, these AI systems could significantly improve healthcare accessibility, awareness, and informed decision-making across Africa.
About the Organizers
The challenge is spearheaded by the Hub for Artificial Intelligence in Maternal , Sexual and Reproductive Health ( HASH ), a multidisciplinary consortium bringing together experts in healthcare, public health, artificial intelligence, data science, and social sciences.
The consortium includes:
- Infectious Disease Institute (IDI)
- Makerere University
- Makerere Centre for Artificial Intelligence (MAK-AI)
- Sunbird AI
HASH focuses on advancing AI-driven healthcare solutions across Sub-Saharan Africa, particularly in areas such as:
- Maternal health
- HIV care
- Sexually transmitted infections
- Adolescent sexual and reproductive health
Their goal is to promote locally grounded AI innovations tailored specifically to African realities and healthcare systems.
Challenge Objective
Participants are required to build a multilingual AI model capable of:
- Understanding health-related questions posed in low-resource African languages
- Generating accurate and contextually appropriate answers
- Responding fluently in the same language as the question
The challenge dataset consists of curated health question-and-answer pairs focused on maternal, sexual, and reproductive health.
Competitors are encouraged to use open-source tools, publicly available pretrained models, and freely accessible datasets where applicable.
Supported Skills and Technologies
This challenge is especially suitable for individuals interested in:
- Large Language Models (LLMs)
- Natural Language Processing (NLP)
- Machine Translation
- Multilingual AI
- Low-resource language modeling
- Deep learning
- Healthcare AI systems
Participants will gain hands-on experience building socially impactful AI technologies for African communities.
Prize Pool
The challenge offers the following prizes:
- 1st Prize — $2,500 USD
- 2nd Prize — $1,500 USD
- 3rd Prize — $1,000 USD
Additionally, participants can earn:
- 5,000 Zindi points
- Recognition within the African AI ecosystem
- Opportunities for collaboration and visibility in AI for healthcare
Evaluation Metrics
Submissions will be evaluated using a multi-metric scoring system combining lexical overlap, structural similarity, and semantic judgment.
ROUGE-1 F1 (37%)
Measures overlap of individual words between prediction and reference answers.
ROUGE-L F1 (37%)
Measures sentence structure similarity using the Longest Common Subsequence (LCS).
LLM-as-a-Judge (26%)
Uses a language model to evaluate:
- Factual accuracy
- Completeness
- Language appropriateness
- Overall response quality
Top-performing solutions will also undergo additional evaluation using AfroLM BertScore F1, which measures semantic similarity using embeddings from AfroLM : a multilingual transformer pretrained on 23 African languages.
Important Rules and Requirements
Participants should note the following:
- Maximum team size: 4 people
- Fully online participation
- Open-source tools only
- No paid services or proprietary tools allowed
- Automated machine learning tools are prohibited
- Maximum of 5 submissions per day
- Maximum of 50 submissions overall
- Top 10 teams must submit reproducible code and reports
- Final winners must publicly share their code and models
The challenge also enforces strict anti-cheating policies, including penalties and disqualification for multiple accounts, unauthorized collaboration, or unreproducible submissions.
Webinar Information
Participants are encouraged to register for the official webinar scheduled for:
Date: 20 May 2026
Time: 6:00 PM – 7:00 PM GMT+2
The webinar will likely provide guidance on:
- Dataset structure
- Evaluation methods
- Submission process
- Baseline models
- Best practices for multilingual healthcare AI systems
Deadline
Application Deadline: 21 July 2026
Why You Should Participate
This challenge is more than a competition. It is an opportunity to contribute to the future of inclusive healthcare technology in Africa.
By participating, you can:
- Build real-world AI solutions with social impact
- Strengthen your NLP and LLM expertise
- Gain exposure within the African AI ecosystem
- Collaborate with innovators across the continent
- Contribute to healthcare accessibility in underserved communities
- Compete for international recognition and cash prizes
For students, researchers, startups, healthcare innovators, and machine learning engineers passionate about African language technologies, this challenge represents a rare opportunity to combine technical innovation with meaningful impact.
Deadline
21st July ,2026.
For more information on this opportunity, click here.
For more global opportunities, click here.
