Engaging live tutoring empowered by AI for all to foster student efficient learning

We are tackling the critical issue of educational disparities, highlighted by the widening performance gap among students in OECD countries, exemplified by France's deteriorating scores in the PISA 2022 assessment. In France, the gap between the most and least privileged students in mathematics is 113 points at the age of 15—approximately 3.5 years of schooling. This stark inequality not only impacts individual student outcomes but also threatens broader societal structures.

Professor Pierre-Michel Menger articulates the urgency of addressing these disparities: "It is urgent for democracies to work on this shape as inequalities in performance engender risks for democratic systems." This warning is pivotal as we consider the implications of an undereducated populace in a world increasingly governed by complex technologies and AI. These inequalities forecast not only future disparities in employment and wages but also a potential democratic crisis, with a generation unequipped to engage with or challenge the technological frameworks shaping their lives.

Our initiative aims to directly address and mitigate these educational inequalities through targeted interventions that enhance learning outcomes for underperforming students, thereby supporting a more equitable and democratic future.

To support students in their learning journey and improve their engagement, attention and learning outcomes, MASTEUR has built a live tutoring app that addresses the root causes of learning poverty through 5 ingredients:

• Data enriched tutoring. Tutoring is provided by top students from top ranking universities and schools, supervised in real time using data on their students' engagement (productivity, attendance, and interaction).
• Scientifically backed method. An evidence-based, sequenced pedagogy that enhances students' focus and learning skills.
• Tailor made learning process by AI. AI technology is not used to provide outputs in a silver platter, which shortcuts and inhibits learning. AI designs customized and multimodal paths in accordance with students' homework and knowledge to improve learning process efficiency.
• Mutliplayer gaming system. A gaming system that promotes cooperation and competition among students based on their cognitive effort in order to engage student in a task flow from knowledge understanding to memorization and consolidation.
• Social driven reward system. Students earn virtual credits as rewards, which they can use to fund meal donations through our NPO partners. Solidarity is used as a leverage to stimulate students’ extrinsic motivation and regularity.

The evidence based Masteur’s method leverages data enriched tutoring with an engaging app, to enhance students’ learning skills. The app measures cognitive effort using photos of handwritten assignments or audio recordings. AI analysis then integrates this effort into a personalized and gamified "flow" for concept assimilation. As students study, they can compete, collaborate with each other, and finance positive actions such as meal donations and tree planting. Masteur offers a new, engaging, and stimulating study experience for all secondary students.

Our solution targets middle and high school students who are underperforming academically, a challenge that affects 25-30% of secondary students across OECD countries. These students often come from less privileged backgrounds and are increasingly distracted, with teenagers spending an average of over four hours a day on smartphones for entertainment. This population is underserved in traditional educational systems, which fail to engage them effectively or address their unique learning needs.

Our program leverages science, technology and human to offer personalized, affordable tutoring services—typically an expensive resource that exacerbates educational inequalities. By making such services accessible, we aim to:

• Improve Academic Outcomes: Enhance performance in fundamentals, thereby reducing the educational gap and offering these students better prospects for higher education and employment.
• Increase Engagement: Create a more engaging digital learning environment that competes with digital distractions by using interactive and adaptive learning tools to appeal the overly connected youth.
• Promote Equity in Education: Provide high-quality educational support to those who typically cannot afford it, helping to level the playing field and reduce long-term societal inequalities.

In brief, our solution not only boosts academic performance but also addresses the broader socio-economic barriers that these students face, preparing them for a successful future in an increasingly digital world.

We are a geek team passionate by Education. Both founders grew up in North African Imigrant families at Paris's underprivileged suburbs, where they experienced the educational challenges firsthand. Their encounters with mentors have been precious. They gain entry into top French engineering schools exemplifying the potential impact of targeted support and guidance.

This personal background not only inspires our mission but also brings an authentic understanding of the socio-economic barriers that our students face. Before launching Masteur both founders interrupted Consultancy carreer, to have 14 month field experience researching, teaching and testing alternative pedagogic layouts with more than 500 secondary students.

Our team extends this insight by including experts like Pauline Roveyaz, a specialist in cognitive science, and tech leaders, alongside high-level mentors such as Safia Tmiri, a former partner at McKinsey MENA, leading Education practice, who brings strategic oversight to our initiatives.

We engage deeply with the communities we serve, drawing on our founders' experiences and our experts' knowledge to ensure that our solution is both effective and culturally resonant. The design and implementation of our program are continuously shaped by direct feedback from these communities.

We chose this stage because our platform has successfully transitioned from concept to pilot testing, demonstrating tangible results and positive user engagement. Over the past 20 months, including 14 months of field exploration and 6 months of rigorous testing (+500 beta users), we have developed and refined a Minimum Viable Product (MVP) that addresses the educational challenges faced by overly connected students.

Our pilot has engaged 1,000 users, including 100 subscribers and 35 tutors, who are actively using our platform. The feedback has been overwhelmingly positive, with users reporting significant improvements in both learning outcomes and engagement. 

In terms of KPIs:

- Engagement score (measured based on various proxies) increase by 10 points in 2 months

- Additional daily studying time per user increases by 24 min during school days. 

- 4.9/5 average satisfaction rate score by >350 users

This validation from our community confirms that our solution is not only needed but is also effective and well-received by those it aims to help.

We are applying to Solve because we believe it offers a unique ecosystem of support necessary to scale our educational platform and increase its impact. 

Firstly, our primary interest in Solve stems from its capacity to facilitate valuable partnerships and provide access to a broad network of expertise in technology, education, and social impact.

Besides, multicultural backgrounds of the team, our pilot is rooted in France field and context. We believe that our solution is universal because of it is scientifically backed and because we are able to easily absorb local context specificities thanks to our tutors that are recruited directly from the communities we target.   

In addition to that, we face technical challenges in refining our AI-driven learning tools to ensure they are as effective and user-friendly as possible across diverse learning environments. We think that we can enrich our product development thanks to SOLVE programme. 

Besides, we encounter market barriers in reaching more schools and communities, particularly in navigating the complex landscape of educational policy and funding within different regions. 

Solve’s mentorship opportunities and strategic partnerships could help us overcome these barriers by connecting us with tech leaders who can provide insights into advanced AI applications and with educational experts who understand the intricacies of deploying new technologies in schools. Moreover, Solve’s community could help us navigate the legal and cultural challenges of expanding internationally, ensuring our solution is culturally adapted and legally compliant across borders.

With our current pilot and its promising results, we think it is the best time for our project to beneficiate from SOLVE, to scale it and convert thousand of tutors into as many solvers. 

Masteur incorporate AI in more parts of the tutoring journey through:

• Automatized tutoring guides and problem sets - train the model to generate the tutoring guides and problem-sets through partnership with existing curricula (public and private)
• Personalized learning processes for students – through better baselining of student starting point, learning trajectory across our key metrics and projected end state.
• Support to tutors - especially in menial activities such as grading, preparing the session, feedback etc.
• Engage students positively - our game based pedagogic system values all student cognitive effort into virtual credits that they could use to finance positive actions such as meal donations. 

Masteur app is designed to close the achievement gap for underprivileged students by providing personalized, technology-enhanced tutoring services.

The theory of change for our solution can be outlined as follows:

Activities: We develop and deploy an AI-driven platform that provides adaptive tutoring guidelines based on student homeworks (pictured). These modules are tailored to individual learning styles and paces, incorporating immediate feedback (thanks to an OCR and generative AI technology).

Immediate Outputs:

1. Increased student engagement due to interactive and personalized content. We measure and follow an engagement score based on proxies (productivity, attendancy, and interactions - more than 40 criteria)
2. Improved understanding and feeling of progress (based on self assessment).

Short-term Outcomes:

1. Students demonstrate a measurable improvement in their overall performance (PISA results in science, maths and reading) and school exams.
2. Increased confidence and motivation among students, as they gain mastery over subjects that previously challenged them.

Long-term Outcomes:

1. Students are better prepared for higher education and competitive exams, opening up further educational and career opportunities.
2. Reduction in educational inequalities as students from disadvantaged backgrounds achieve academic success comparable to their more privileged peers.

Evidence Supporting Links:

• A recent research based on metanalysis has proven tutoring efficiency. On average, tutoring increased achievement by roughly an additional 3 to 15 months of learning across grade levels (K12) (source: https://annenberg.brown.edu/sites/default/files/EdResearch_for_Recovery_Design_Principles_1.pdf).
• Our pilot testing have showed an additional daily studying time per user by 24 minutes (measured by the app), validating the effectiveness of our solution in regards to student engagement.
• Feedback from students and teachers during our initial rollout confirms increased engagement and satisfaction with the learning process (4.9/5 rating based on 350 student feedbacks).
• Our validation process are designed and backed by our cognitive scientist, Pauline Roveyaz, who is a farmer collobarator of Stanislas Dehaene. 

By systematically linking our activities to these outputs and outcomes, our platform is set to not only improve student engagement and academic performance but also address broader educational disparities, contributing to a more equitable society.

Our primary impact goal is to significantly improve educational outcomes for underprivileged students, particularly in mathematics, thereby reducing the academic performance gap between socio-economic groups. This aligns with the UN Sustainable Development Goal 4, which aims to ensure inclusive and equitable quality education and promote lifelong learning opportunities for all.

1. Increase Mathematics Proficiency: We aim for at least a 30% improvement in mathematics proficiency among our users (1 year of schooling).

2. Boost Student Engagement and Confidence: Our goal is to enhance engagement and academic confidence

3. In longer term, Reduce Educational Inequalities thanks data insights on student engagement: We seek to decrease the disparity in educational attainment between different socio-economic groups by:

   • Demographic Performance Analysis: Assessing improvements across various demographics to ensure equitable impact.
   • School Performance Data: Collaborating with schools to track overall changes in school-wide performance metrics.

Progress towards these goals is monitored through a combination of in-platform analytics, periodic academic assessments, scientifically backed validation processes, and feedback mechanisms that involve direct input from students and tutors. 

Our solution is based on AI technology and incorporates it throughout the learning journey.

• Data enriched tutoring. Tutoring is provided by top students from top ranking universities and schools, supervised in real time using data on their students' engagement (productivity, attendance, and interaction).
• Scientifically backed method. An evidence-based, sequenced pedagogy that enhances students' focus and learning skills.
• Tailor made learning process by AI. AI technology is not used to provide outputs in a silver platter, which shortcuts and inhibits learning. AI designs customized and multimodal paths in accordance with students' homework and knowledge to improve learning process efficiency.
• Mutliplayer gaming system. A gaming system that promotes cooperation and competition among students based on their cognitive effort in order to engage student in a task flow from knowledge understanding to memorization and consolidation.

Fida HAMROUNI, cofounder and CEO, Mines Paristech 2013, Ex EY

Wajih JMAIEL, cofounder and CTO, Telecom Paristech 2015, Ex Wavestone

2 Tech + 1 part time cog. science expert (Pauline Roveyaz, ENS) + tutoring manager + 40 tutors students in top France Universities

Before launching Masteur, founders have created a non profit organization that allowed during 8 years to mentor more than 200 secondary students with poor background, enriching field knowledge of the founder team. 

In addition to that, we have spent over 16 months developing and testing a product that is loved by more than 1000 students and approved by their parents.

• Recruitment: We implement bias-aware recruitment strategies to attract diverse candidates, ensuring a wide array of perspectives within our team (academical backgrounds, gender, ethnical origins and social background)
• Professional Development: We provide mentorship and advancement opportunities particularly for underrepresented staff, promoting diversity in leadership roles (students from high education comming from poor background and trained and employer to inspire next generation). 

Except platform expenses, costs are limited to tutors' remuneration. Today it reaches 25€/h. As tutors follows between 3 to 5 students per 1 hour live session, we are able to drastically reduced costs to parents. Today, parents pay 13.5€/h, which is quite competitive. We want to decrease these costs to reach 5€/h and develop Public-Private Partnerships in order to make Masteur affordable and accessible for all. 

We achieved a 6-month traction in April from 1k€ Monthly Recurring Revenue in october to 12k€ in April with a 27% MoM MRR growth and 25% margin growth.