Ghana, and to a large extent, Africa, is facing a lot of challenges in the area of education, as many institutions lack the necessary and appropriate technologies to ensure quality teaching and learning.
George Jojo Boateng, a computer scientist, engineer and social entrepreneur, is hoping to bring an end to these challenges in the educational sector through AI-powered tools and apps.
Currently pursuing his PhD in Applied Machine Learning at ETH Zurich, Switzerland, George is also the President and co-founder of Nsesa Foundation which runs SuaCode.
According to him, SuaCode is the most exciting project he has worked on.
SuaCode is a smartphone-based online coding course which teaches millions people across Africa on how to code using smartphones.
“From 2018 to 2020, with 4 pilots of SuaCode, we’ve had 3K applicants from 69 countries, accepted 1K students, with a 62% completion rate and alums getting internships, jobs and studying computer science and engineering in top schools e.g., Yale, Columbia, Dartmouth, MIT etc. The potential of SuaCode is just huge!” George told Kuulpeeps.
To enhance teaching and learning on SuaCode, George built a bilingual AI teaching assistant called Kwame, to provide quick and accurate answers to students’ coding questions in English and French.
Explaining how the idea for Kwame AI teaching assistant was birthed, George said:
“In 2017, we developed a smartphone-based coding course in Ghana, SuaCode, that enabled students to learn to code on their phones circumventing the need for computers. We launched an online version of SuaCode in 2018 and expanded it beyond Ghana to all Africans, dubbed SuaCode Africa in 2019. For our most recent cohort that run from May to June 2020 — SuaCode Africa 2.0, the course was offered in both English and French.”
“Over 2,300 students across 69 countries, 42 of which are in Africa applied. We accepted and trained 740 students. With our students needing a lot of assistance in their first coding course ever, we have relied on human facilitators to provide support and answer students’ questions. For example, in SuaCode Africa 2.0, facilitators contributed over 1,000 hours of assistance time for an 8-week period and helped to achieve an average response time of 6 minutes throughout the course.”
“This approach is, however, not scalable as the number of students applying to SuaCode is increasing exponentially. I saw a potential to solve this problem with AI, so I built Kwame to do just that.”
“I took a course at ETH Zurich from February to June 2020 where for my course project, I built Kwame and wrote a paper on the work which I later presented in a workshop at NeurIPS 2020 (a top AI conference) and recently at the 22nd International Conference on Artificial Intelligence in Education.”
George further revealed that the AI teaching assistant was named, ‘Kwame’, after Ghana’s first president Dr Kwame Nkrumah by his co-founder.
“Kwame is named after Dr. Kwame Nkrumah, the first President of Ghana and a Pan Africanist whose vision for a developed and empowered Africa resonates with the goals of SuaCode. My cofounder, Victor Kumbol actually came up with the name. He has a knack for coming up with cool and meaningful names.”
George, however, believes that the Kwame AI teaching assistant will revolutionize remote education across Africa.
Explaining why, he said: “Just imagine a primary school student in a rural part of Ghana calling a toll-free number using a feature phone and asking about photosynthesis in Twi (a Ghanaian language), and then getting a response from Kwame explaining photosynthesis in Twi. And the interaction goes on for like 1 hour with more question answering. During the interaction, Kwame after giving a certain answer detects frustration or confusion from the student’s voice and reply. Then Kwame proceeds to adapts his answer.”
“This is the kind of future my team and I are building! One in which a student, in any part of the continent (e.g. rural area), with the most easily accessible technology to them (e.g. feature phone) can still have quality education (get concepts clarified/explained to them anyplace, anytime, for an unlimited duration). And that is why I believe Kwame will revolutionize remote education across Africa,” he said.
Aside from the Kwame AI teaching assistant, George has also built other AI-powered apps including apps for stress detection (whether someone is feeling stressed or not), physical activity detection (whether someone is sitting, walking or running), voice activity detection (whether someone is speaking or not).
He also revealed that for his PhD, he is building a smartwatch app to recognize the emotions of romantic couples managing a chronic disease based on their conversations and interactions.
George hopes that the Kwame AI teaching assistant will become bigger in the near future, to the extent that it will be used globally to enhance education.
“We want to shape a world where young Africans grow up learning coding skills with their phones and our Kwame AI, and they believing that they can create the next big technological breakthrough!”
“Our immediate next step is the deployment of Kwame in real-world courses. To achieve that, we’ve built a platform, SuaCode.ai and we will be beta testing it in the next few months. People who are interested in experiencing Kwame can sign up here: bit.ly/sc-beta-test“
“Furthermore, we’ve been working on Kwame-as-a-Service (KaaS) so that various education or EdTech companies across Africa can use our AI teaching assistant in their courses via an API or alternatively, our SuaCode.ai platform. We are looking to partner and so interested organizations or companies can fill this form: bit.ly/sc-kaas“
He continued: “In five years, we expect to have millions of people across all 54 African countries using our SuaCode.ai platform monthly, taking courses, interacting with Kwame in African languages, getting internships and jobs. We would have built ground-breaking AI systems validated in the real-world for the African context. Our impact would be so tangible that when anyone thinks of Education + AI, they would think of SuaCode.ai!”
“Our SuaCode courses target people between the ages of 13 and 35 (so secondary school and tertiary), mainly because of our expected minimum reading level and cognitive skills. But anyone really can take our courses. If anyone is interested in taking our SuaCode course, they can sign up here for our upcoming cohort: bit.ly/sc-beta-test,” he added.
Click on the comment box below and leave us your thoughts. Thank you