5 Sep 2018, 14:51

Interview with Weever Mbakaya, co-founder of AI Kenya

Weever is an avid learner and technologist who is passionate about growing artificial intelligence in Africa. He has a background in computer science with a specialization in applied machine learning.

Tell us how you got involved with data science and AI

I have been a serial Python software engineer for a while and I was looking for a better way of solving some of the most significant problems I was encountering. There was a natural progression leading into me undertaking a Masters with a specialization in machine learning while learning data science through online courses.

What are your favourite machine learning libraries?


Tensorflow would be 1st on my list. It is easy to work with on a production project and implements parallelism in its computation which gives a huge advantage in speed of developing an AI solution.


Because it is comprehensive but not overwhelming and can help beginners in approaches to solving problems. It’s also great for production when resources and time are limited


I actively use deep learning methods and its the best at this & allows using combinations of other libraries like Blocks for deep learning projects.

What led to you starting AI Kenya?

AI Kenya started out of the need to tackle the challenges in my country that could be solved through data science and machine learning - yet very few people practicing the technology and businesses not actively adopting the technology.

Where would you like to see AI Kenya in 5 years time?

For it to be among the hubs of AI in Africa with research projects available for interested practitioners to work on and build AI solutions for Africa, in Africa.

I’d like to see AI Kenya contributing to the success of many individuals in becoming domain experts, actively impacting the technology industry in Africa.

What are the biggest areas of opportunity you see in African AI over the next few years?

1. Health.

There are many opportunities for African startups that will come out of AI technology to solve health issues. For example diagnosis AI, recommendation bots and improving response to health emergencies.

2. Finance

This industry is already slowly adopting AI in existing businesses and startups are on the hunt of data-driven credit scoring methods, seamless payment across different platforms and fraud detection.

3. Agriculture

This will include exploring methods of practicing agriculture at a cheaper cost and guaranteeing a market for it. This would possibly include drone & smart farming methods fusing AI and IoT devices.

What advice would you give Zindi competitors for solving competitions?

Two things:

1. Understand what you’re solving

Understand the data used in the competition and focus on feature engineering to help you understand and create a strategy to generate the best solution for the competition

2. Choose wisely and test often

Choose the right method and algorithm to give the best accuracy and always cross validate

For more on AI Kenya, join a Meet Up or join their Facebook Group