Liberating Education (LE) is a nonprofit organization located in Karlsruhe, Germany, focused on expanding higher education and entrepreneurship training to underserved communities. Its hybrid learning platform currently serves 149 registered students across eleven learning centers in Africa and Asia.

As the platform continues to expand into new communities, learners need a way to see what skills they have learned from the lessons offered, and where gaps in their skills may exist. Additionally, students need exciting new ways to stay engaged with their learning.

Our LE AI Learning OS expands the LE learning platform by enabling students to visualize their learning through two graphs: a knowledge graph and a soft-skills graph. The knowledge graph contains nodes with information about topics a student has learned from lessons. The soft-skills graph appears as a radar chart with ten skills a learner can improve throughout their time on the platform.

Students improve their scores on the skill graphs through interactive activities such as flashcards, matching, and hexagonal thinking. Flashcards promote prompt-based learning on questions from various course topics. Matching activities provide definition-matching quizzes for different lessons. Hexagonal thinking focuses on building descriptive connections between topics. The knowledge and skill graphs are updated based on these activities to provide insights into student performance.

Our software improves the existing LE learning platform by providing engaging activities and tracking learning outcomes, furthering the goal of making education more accessible.

Our software uses Next.js for the front end. The back end uses PostgreSQL and Supabase for data storage and uses Anthropic’s Claude API for flashcard generation and grading. Our back end connects to the front end using FastAPI.