Skip to content

HudsonGri/llm-os-agent

Repository files navigation

OS LLM Agent

An AI-powered learning assistant for Operating Systems students that provides contextual help using course materials and tracks learning patterns to improve course delivery.

Overview

The OS LLM Agent is a specialized chatbot designed to enhance the learning experience in the Operating System (COP4600) course at The University of Florida. By leveraging Large Language Models and Retrieval Augmented Generation (RAG), it provides students with 24/7 access to course-specific assistance while giving instructors valuable insights into student learning patterns.

Key Features

  • Contextual Course Assistance: Get immediate, accurate answers based on official course materials
  • Resource References: Direct links to relevant slides and course materials in responses
  • Canvas Integration: Seamless access through Canvas course pages
  • Analytics Dashboard: Insights for instructors on common student questions and pain points
  • Continuous Learning: System improves over time through usage patterns and feedback

Tech Stack

  • Frontend: Next.js, React, Tailwind CSS, Shadcn UI
  • Backend: Python, Next.js API Routes
  • Database: PostgreSQL with pgvector for embeddings
  • AI/ML: OpenAI/Google LLMs, text-embedding-3 for RAG
  • Integration: Canvas LMS API

Team

  • Michael Tang
  • Hudson Griffith
  • Erik Hartker

Acknowledgments

Project developed under the guidance of Alexandre Gomes de Siqueira at the University of Florida.

About

LM Agent for Operating Systems Course Assistance

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •