Welcome to the official GitHub organization for Team USUTVKU. We are a group of passionate students from VNU-HCMUS, HCMUT, and VKU, working together to redefine the digital learning experience for the GDGoC Hackathon 2026.

Our project is a Customizable Duolingo-style Learning Platform designed to solve the "Passive Learning Trap." By integrating advanced AI agents and Graph RAG technology, we move beyond simple streak retention to focus on deep cognitive processing and structured knowledge application.

• Deep Processing: Moving students from passive watching to active concept application.
• Socratic AI: Our agents provide hints and guidance rather than direct answers to prevent the "AI Crutch" problem.
• Knowledge Mapping: Utilizing Graph RAG to visualize how concepts (like Python variables) relate to broader topics (like Data Structures).
• FSRS Spaced Repetition: Optimized retention using modern flashcard algorithms.
• Seamless Note Taking: Support taking notes in markdown directly on the platform and let AI agents ingest them.

This organization hosts the two core components of our platform:

The powerhouse of the platform, handling complex AI workflows and data persistence.

• Stack: Python (FastAPI), PostgreSQL, Redis, Neo4j, Milvus.
• Key Features: Orchestrated AI Agents, Graph RAG implementation, Spaced Repetition (FSRS) service, and Celery-based task processing.

A high-performance, interactive interface built for a seamless user experience.

• Stack: Next.js 15, TypeScript, Tailwind CSS, TanStack Query.
• Key Features: Interactive Knowledge Graph visualizer, Markdown-based notes editor, and real-time gamification dashboards.

We are building the future of personalized education. If you're interested in our research on Graph RAG or FSRS optimization, feel free to explore our repositories or reach out to the team members.