MathIAfrica

Although more children than ever are enrolled in school, far too many are not learning. In francophone Subsaharan Africa hundreds of millions of students live in “learning poverty”- that is, they are either out of school or far from reaching basic proficiency in literacy and numeracy (e.g. 97% of school-aged children in Chad, 86% of school-aged children in DRC, 82% of school-aged children in Ivory Coast are “learning poor”, according to the World Bank). COVID-19-related closures have worsened this scenario: school closures impacted around 250 million students in sub-Saharan Africa: learning completely stopped for most of them. 

Barriers to learning are multifaceted and range from financial barriers to suboptimal public expenditure in education, subpar teacher training and weak incentives for teachers to perform, alongside outdated curricula and pedagogical approaches. While these require systemic change, we believe that AI-enabled learning systems could play a role in enabling an improved learning experience, increased tailoring and flexibility in teaching practice, the provision of timely feedback, flexibility in managing students’ learning experiences and faster student progression.

Pedagogical innovations such as technology-assisted instruction, remedial education, and tracking by achievement have proven potential to improve learning at a low cost, when delivered at scale. Our solution, MathIAfrica, is based on Mathia, an award winning EdTech solution successfully deployed in France starting 2020 by Prof en Poche. Mathia is winner of the Artificial Intelligence Innovation Partnership (P2IA) for teaching mathematics in France, and the first public contract in artificial intelligence from the Ministry of National Education, Youth and Sports in France. Mathia is currently deployed in several academies: Paris, Bordeaux, Toulouse, Créteil, Rennes, Reunion, Nancy-Metz, Lille and Poitiers. 

Mathia provides an AI-enabled, gamified experience to learn mathematics and sciences. Its main added value is that it makes the learning experience immersive and engaging, thanks to a combination of voice interaction, 3D visualizations and adaptive gamification. Mathia is designed for all, including students with special needs, online and offline. Mathia leverages AI to adapt the level of the questions to the level of each student. Its gamified approach allows the student to become an actor in an educational game in which she is the protagonist and on which she progresses at her own pace. 

With MathIAfrica, we look forward to adapting Mathia to low-resource, low-connectivity settings, in which it could offer support to teachers and students.

Several national mathematics curricula in many countries of Francophone Africa are modeled on the French curriculum, and are delivered in French. These similarities can be leveraged to adapt existing high-quality learning resources to the context of several Francophone countries characterized by high levels of learning poverty and post-COVID-19 learning losses. We propose to adapt MathIA into MathIAfrica to serve children who are falling behind in mathematics and science in francophone Africa. The software is designed to offer teachers the possibility of forming groups to implement “teaching at the right level” (TaRL), gives teachers summary information on the acquisition of key concepts and skills and allows teachers to manage the classroom in evidence-based fashion. TaRL works by dividing children into groups based on learning needs rather than age or grade; dedicating time to basic skills rather than focusing solely on the curriculum; and regularly assessing student performance.

Mathia is flexible and can be used for remedial education as well as mainstreamed to deliver the national curriculum. Mathia adapts to different scenarios: individually, in class or at home, or in small groups of students; it can work directly from Mathia's website or via mobile application. The great majority of the functionalities can be used offline.

MathIAfrica leverages MathIA to help education actors increase the quality of learning in mathematics in francophone Africa, including those typically excluded from online learning or from accessing innovative education solutions.

EdTech programmes must be designed with learning outcomes as the main goal, since technology in isolation changes little. Reaching all learners requires context-specific and culturally responsive approaches. The partnership between PeP and Principle 7 brings together cutting-edge expertise in EdTech delivery [PeP] with solid experience in the design of education programs in low-resource settings [Principle 7]. 

We regard incremental co-creation with the users - in this case, the teachers - as a fundamental principle for design. To incrementally understand the needs of teachers and learners, we plan to run different small scale pilots in different environments, and co-design with teachers and communities following an AGILE methodology. MathIAfrica must be aligned to national curricula and be of use under local circumstances; therefore it is essential to iteratively pilot it and co-design with national actors to attain the best possible fit within the national education ecosystem. Following this logic, cooperation with the local Ministry of Education, local Universities and national research institutions active in the education field would be particularly strategic. Our partnership has identified the following steps to develop a demo for Cote d’Ivoire and / or Senegal, along the following indicative steps:

1. Review of challenges faced by existing in country for online learning in Francophone Africa 

2. ‘Localisation’ of content and delivery modalities - adapting existing package to local users and teachers 

3. Pilot teacher training module iteratively  

4. Roll out a prototype under different delivery modalities, including via hub devices and offline learning and in different schools [urban setting, rural setting, peri-urban setting, potentially also refugee camp, if feasible]

To join a network of impact-minded leaders in the education field who could be able to give feedback / trusted support / strategic guidance.

To access mentorship and strategic advice from experts

To gain exposure in the media and at thematic events, such as fairs and conferences.

The main innovative elements can be described as follows: 

1) making the learning experience immersive and engaging. Thanks to a unique combination of voice interaction, 3D visualization and artificial intelligence MathIAfrica would make mathematics easier to understand.

2) Inclusive design. MathIAfrica is designed for all, including students with special needs, and offers 3 response options: oral, written (drawing) or numeric keyboard. 

3) An additional innovative aspect is that the software explicitly aims at creating positive emotions in children through a tailored feedback mechanism that involves voice recordings, too. Through experiments in the classroom and at home, many teachers expressed positive feedback on MathIA. For example, Aurelio, Teacher at the European School of Varese says “My students have established a real emotional relationship with the friendly little robot who follows them, encourages them while adapting to their rhythm, their strengths and weaknesses.”

Another example: Rallou, Teacher at the European School of Brussels: "This solution allows students to discover, understand and learn important math concepts in a fun and engaging way”.

We expect that MathIAfrica may also stimulate market demand for similar learning solutions, increase capacity of Ministries / schools / education stakeholders to deliver EdTech, and enable broader positive impacts via generation of lessons learnt.

MathIAfrica pursues the following goals:

1. Improving learning outcomes in primary school in mathematics in low-resource settings.

2. Capacity development of teachers: a) provide teachers with educational assistance for the differentiation and personalization of learning through innovative solutions based on scientific research; b) Enable a growing number of teachers to facilitate learning for different styles, and at different paces. Any introduction of personalized learning technology should not be interpreted as decreasing the importance of the teacher, but rather enhancing it. 

3. Develop flexible learning resources usable in delivery of either formal education, remedial education and /or accelerated learning programmes delivering “compressed” curricula for children who have missed out on long periods of schooling.

SMART Indicators for key outcomes would be:

Number and % of children who use MathIAfrica achieving minimum proficiency in mathematics, as defined by national standards or by standards available in the global proficiency framework.

Number / % of teachers who, after being trained, successfully implement software-supported TaRL principles in the classroom

Number / % of teachers who, after being trained, express satisfaction with the learning solution

Should we be able to mobilize a sufficient budget, we would have the technical capacity to compare skills acquisition in different schools, tracking students’ skills acquisition with and without access to our learning solution.

Indicators would be disaggregated to the maximum possible extent by sex, location, age, background and other relevant demographics, to understand how different subgroups are impacted.

If teachers are trained on leveraging high-end Edtech solutions, and

If technocrats from Ministries of education are able to quantify the benefits of high-end Edtech solutions, and

If parents see the benefits of innovative pedagogy (AI-enabled TaRL) in terms of learning outcomes

then

• Children are more likely to catch up on learning losses,

• Teachers are better equipped to delivery high-quality content, potentially to a larger number of children and at a low per student cost

• Technocrats are more likely to endorse / understand the benefits of Ed-Tech solutions that could help boost teachers’ productivity and be cost-effective at scale

• Practitioners and researchers could benefit from collective learning on the conditions in which AI adaptive solutions can deliver learning in low resource settings

Our solution mainly relies on AI, but could be integrated with locally relevant low-tech solutions; we would have more elements once we engage in the co-design pilots with local teachers (as described above) to adapt the "mother" application (MathIA) into MathIAfrica.

About 40% of the team members are women. Two developers come from priority urbanization areas and one web developer is a person with a disability. We come from several countries and backgrounds; the team has several years' worth of work experience in developing countries.

Our team is passionate and very curious about how our world works. 100% of our team wants to pass on this passion to children, and in particular to those with special educational needs. So, for example, our platform is adapted for children with dyslexia, and work is underway to make the platform adaptable for users with color blindness.

Our main value proposition is to help both the teacher to teach and the student to learn. Thanks to the AI, we empower the teacher by giving instant feedback and smart recommendations for a more personnalized teaching approach. We also use AI to interact in the most gamified way with the student, making the process of learning mathematics much more fun and easy.

Our current business model consists in charging a yearly fee for teachers using Mathia.

Usually, the clients can be : Ministries of education (such as the French Ministry of National Education, our main client), regions and districts, and of course schools themselves.

Prof en Poche has been an innovative French EdTech editor for more than 5 years. We currently have a contract with the French government until 2026 and we are deploying our solution in french speaking schools abroad (Belgium, United Kingdom, Morocco). We also launched recently the B2C version of Mathia.

The company has always been profitable since 2017.

Our work has been saluted by different prizes :

- Laureate of the i-Nov (BPI France) contest in 2021 : 500K€ in non-equity funding
- Laureate of the IMPACT EdTech Program (European Commission) in 2020/2021 : 140K€ in non-equity funding

More than 1100 families subscribed to our service (B2C) in less than a year and Mathia is deployed to thousands of schools all over France (B2B).