Angaza Elimu: AI powered adaptive and interactive learning platform
The UNICEF Innovation Fund is proud to see a portfolio member Angaza Elimu graduate. They’ve come a long way – from numerous product iterations to deep diving into understanding their ecosystem better and strengthening their business model. They’re now ready to collaborate at a larger scale – as they find new pathways to work with partners, investors, and the open source community.
Over the past 12 months, our team was able to build a state of the art educational data mining model with the goal of powering intelligent learning and teaching interventions. Through our learning platform students receive a more personalized experience allowing students to learn at their own pace and promote self development. . Our solution allows teachers access to real time actionable data insights helping them save more time to address students needs and provide solid evidence to support their teaching interventions.
Over the last year, our student users have grown from 9000 to 30,000 users with an average daily user engagement of 29 mins.
During the investment period our team focused on three key product features:
Educational data mining model (83% accuracy):
Adaptive learning: we have built a robust adaptive assessment model for the curriculum concepts. Here students are able to learn and assess themselves at their own pace.
Actionable insights: through the model we are able to derive actionable insights on learning and teaching outcomes. We will be sharing these insights with different education stakeholders with a goal of supporting them to understand key problems, needs and opportunities to invest in.
Offline Interactive mobile app: through this app students are able to continue learning(taking lesson notes, assessments, ask a teacher, peer discussions) even when not connected to the internet ;teaching and learning can be conducted in remote areas with limited access to the internet .
Innovators STEAME web app: through this app students are able to get a personalized learning experience of relevant 21st century skills such as Internet of Things(IoT), robotics, mechatronics, embedded systems, web and mobile app development, game development and animations. Instructor and self paced learning is also available on this web app.
We aim to use this product to transform Angaza Elimu to a fully fledged data driven EdTech company. The main goal is to enable stakeholders in education to make data-driven decisions.
PROTOTYPING & USER TESTING
We had an intense iterative product development life cycle with well defined phases owned by core product team members. The following were the key phases: Product design, Designing the solution, Building the solution, Unit user testing,Validation and feedback.
We did our first test of the model with 547 students in different primary schools in Kenya. The goal was to determine the relationship between the learning engagement process and student performance. Some of the key lessons learnt from this test are:
Relation between student engagement to resources and learning outcomes: students accessing more learning resources like lesson notes and assessments performed better than those who were less engaged.
Relation between teacher-student engagement and learning outcomes: students who engaged their teachers more in seeking further clarification of the curriculum concepts quickly mastered the concepts and recorded a higher improvement in their performance.
We had interesting outliers from the test data where some few students with less engagement recorded good performance. This will help us to keep refining the model.
ON BEING OPEN SOURCE
By engaging in the open source technical assistance program offered by the Fund, we developed great standard documentation and coding practices. With the practices implemented we will be able to easily attract a huge pool of community contributors as they will find it easy to contribute to our project.
Over the last 12 months we have had a huge focus on growing the business through strategic partnerships. The main reason behind this is to expand our reach to more rural areas. We have had great success in increasing our partnership portfolio. Some of the key partners are: Google for Startups(received funding through the Black Founders Fund and support from Google experts) and Liquid Intelligent Technologies(trained students from 22 secondary schools on IoT). We will continue putting more efforts in growing our partnership portfolio over the next year. We are targeting development partners, donors, corporates and government agencies. Last year we recorded slow B2C growth. This could be attributed to the disruption of learning by COVID-19. We have drafted new strategies to change this slow growth narrative which include implementing agency distribution models.
NEXT SET OF GOALS
In the next 5 years, we want the company to expand its reach to more rural areas. To do this in a sustainable way, we will need to partner with strategic partners to facilitate this. These partners range from development organizations, donors, corporates and government agencies. Through the educational data mining model we aim to support their interventions and strategies with a goal of improving the learning and teaching outcomes. Our team is most excited about our growth plan for the next year. We target to grow our users from the current 33k+ to over 150k+. Apart from growing the traction we are committed to ensure that we impact the users by helping the students achieve about 30% improvement on their academic performance.
UNICEF INNOVATION FUND
Working with the Fund has been an incredible experience for our company. Team members from the Fund have gone above and beyond to provide us with critical support over the last 12 months. The mentorship, constructive criticism offered and networks created through the Fund will go a long way in powering our growth.