xiao18825501901-rgb/coursemate-ai

CourseMate AI — RAG course assistant and study agent

What it solves

CourseMate AI is a personalized teaching platform designed to help students manage multiple courses. It provides a centralized system for interacting with course materials, generating study plans, and tracking tasks, replacing fragmented study habits with a structured, AI-driven approach to learning.

How it works

The platform uses a dual-backend architecture to separate knowledge retrieval and task management:

  • RAG API (FastAPI): Handles the ingestion and retrieval of course documents (PDF, DOCX, PPTX, Markdown, and text). It uses a hybrid retrieval system combining SQLite FTS5 and stored embeddings with reciprocal-rank fusion to provide cited answers based on the course content.
  • Agent API (Express): Manages a tool-based loop for user tasks. It provides five specific tools for creating, searching, updating, completing, and deleting Todo tasks, ensuring that AI-generated actions are owner-isolated and secure.
  • Frontend (React/Vite): A responsive interface that allows users to navigate between QA, study plans, and document management, authenticated via Clerk.

Who it’s for

Students and educators who need a structured way to to organize course materials and interact with them using AI, as well as administrators who can manage the course corpus and moderate community-published courses.

Highlights

  • Hybrid Retrieval: Combines keyword search and vector embeddings for more accurate, cited responses.
  • Task-Driven Agent: An AI agent capable of managing a personal todo list through a set of deterministic tools.
  • Course Isolation: Ensures that data and tasks are isolated by course and user, maintaining privacy and security.
  • Course Creation: Supports private-by-default course creation with a secure upload and indexing pipeline.
  • Multi-lingual Support: Includes routing for Chinese and English tutors.

Related

  • Project
    FutureUniant/WorkShadow

    A local-first desktop work log application that uses AI for semantic search, automated report generation, and Q&A over personal work history while keeping data on the user's machine.

  • Project
    jianruntech/geo-score

    The ​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​ ​  ​

  • Project
    Ais1on/CTI-RAG

    A RAG system specialized for Cyber Threat Intelligence that combines vector search and knowledge graphs to enable multi-hop analysis of attack organizations and malware.