itsmorninghao/CogniLoop

智能助教系统,利用大模型根据课程相关文档自动生成题目并且批改

What it solves

CogniLoop is an AI-driven decentralized knowledge learning community designed to provide personalized, adaptive learning experiences. It solves the problem of one-size-fits-all education by allowing users to create their own knowledge bases and receive customized practice questions based on their individual learning progress.

How it works

The system leverages a combination of LLMs, vector databases (pgvector), and orchestration frameworks (LangChain and LangGraph) to manage knowledge and generate content. Users can build knowledge bases, join learning circles, and participate in challenges. The system tracks each user's learning trajectory to continuously optimize the question-generation strategy, ensuring the material is tailored to the learner's specific needs.

Who it’s for

It is primarily designed for students and learners of all backgrounds, including those in remote areas, who want access to high-quality, customized AI-driven educational tools for practice and study.

Highlights

  • Adaptive Learning: Continuously optimizes question strategies based on user learning paths.
  • Personalized Content: Generates custom practice questions from user-defined knowledge bases.
  • Community Features: Supports the creation of learning circles and challenge competitions.
  • Detailed Tech Stack: Built with FastAPI, PostgreSQL/pgvector, LangChain, LangGraph, and React.

Related

  • Project
  • Project
  • Project
  • Project
  • Project