Despite their exclusion from mainstream conferences and gatherings, artificial intelligence and machine learning communities in Africa are growing and focusing on some of the continent’s most pressing issues.

In August 2019, 700 of Africa’s best machine learning researchers came together at Kenyatta University in Nairobi, Kenya for the third annual Deep Learning. It was the largest Indaba yet and we are proud to have supported the organizers as they build a strong and interconnected AI community across the continent.

Deep Learning Indaba 2019 is an annual conference that aims to strengthen African machine learning and artificial intelligence with the vision of African engineers as active shapers and owners of these technological advances. The theme of this year’s Indaba was #SautiYetu, “our voice” in Swahili. We saw the theme reflected throughout the events from keynotes focused on frontier applications of AI in medicine to African perspectives on innovating for global impact.

Hosted this year in Nairobi, Kenya, the Indaba gathered over 700 machine learning researchers from over 30 countries to present innovative solutions to pressing social problems in their communities. Here, meet some of the researchers addressing the most pressing issues from their communities using machine learning—follow their work around the continent with our StoryMap.


Jordan Sichalwe and Francis Chikweto are using machine learning to address the lack of experienced pathologists in Zambia. Their new cervical cancer dataset combines existing datasets from New York University and samples gathered from Zambia to help algorithms and the doctors they inform identify cancerous cells and diagnose cervical cancer early and accurately.


Genet Shanko Dekebo is using machine learning to help midwives identify risky pregnancies that will need advanced care early. With close friends and family working as midwives, Genet is building community-driven data tools that equip community health care providers like herself with the advance warning that could save a life using minimal additional resources.


Rukia Mwifunyi is using data-driven management to optimize electricity access using existing infrastructure in Tanzania. By modeling hour-by-hour electrical load on the power grid, smart grid technology can maximize the number of people getting access instead of technicians making best guesses that often result in the most vulnerable populations being left without power.


Jean Amukwatse is building a low-cost soil analysis and cropping decision system for farmers in her home country of Uganda. Her technology will identify optimal crops for a farmer’s soil to inform management decisions that could increase income and food security for vulnerable rural communities.


Fauste Ndikumana is using machine learning to mitigate the risks of both pest infestation and overuse of pesticides using an integrated and automated system for agricultural regions in Rwanda. Her system will help identify fall armyworm in time to save critical smallholder crops while also minimizing the use of chemicals dangerous to human health, protecting families reliant on their subsistence food crops.


Sam Masikini is using machine learning to build an app that identifies counterfeit banknotes – a technology that could restore trust in the financial systems of Malawi. In a financial ecosystem dependent on small bills, reducing the circulation of counterfeit notes protects small business owners and supports economic stability.


A highlight of the conference for us was the poster sessions. Posters presented original research on cutting edge AI methods, innovative applications, and projects already incorporating AI for real impact.  Prizes were awarded to Kale-ab Tessera for his research on learning compact, general-purpose neural architectures and Ayodele Oladiyi for his work on identifying signals of financial fraud in text messages. The poster session stood out for its spirit of genuine collaboration between researchers from across the continent.

Across the posters, we were most impressed by the work of researchers using new machine learning methods to solve critical problems in their communities. Some even left behind existing careers to learn to code in the hopes of catalyzing critical changes. They often had to overcome significant data availability and computational challenges to do their work—challenges that would have made many others give up or change topics.

The challenges overcome by many participants showcased their commitment to affecting data-driven solutions for their communities. We believe that events like the Indaba are essential to a positive future for AI in the coming decades; we are excited to be supporting a more diverse and inclusive AI community around the globe, and the Indaba is part of our growing body of work. To learn more, visit the Deep Learning Indaba and stay tuned.

group of men looking over mobile phone
Evan Tachovsky, Lead Data Scientist for The Rockefeller Foundation, discusses an app idea with participants at the Indaba.

The post AI Solutions Sourced Locally: Meet the Deep Learning Indaba Attendees appeared first on The Rockefeller Foundation.

Share with cohorts