CaviraOSS/PageLM

What it solves

PageLM is an open-source education platform designed to transform static study materials like PDFs, documents, and lecture notes into interactive learning resources. It automates the creation of study aids to help students and researchers learn more efficiently.

How it works

Users upload documents or lecture recordings, which the platform processes using LLMs and text-to-speech (TTS) engines. It leverages LangChain and LangGraph to orchestrate AI tasks, allowing users to chat with their documents, generate structured notes, create flashcards, and produce AI-generated podcasts. The platform supports a wide range of AI providers, including local models via Ollama, and offers flexible storage options ranging from simple JSON files to vector databases.

Who it’s for

It is built for students, educators, and researchers who want a customizable, self-hosted alternative to proprietary AI study tools for managing and interacting with educational content.

Highlights

• Diverse Learning Tools: Automatically generates quizzes, flashcards, Cornell-style notes, and audio podcasts from uploaded content.
• Extensive Model Support: Compatible with numerous LLM providers (OpenAI, Claude, Gemini, Grok, Ollama) and multiple TTS engines for audio generation.
• Real-time Interaction: Uses WebSocket streaming for responsive chat and content generation.
• Flexible Deployment: Supports local development and Docker-based production deployment.
• Comprehensive Features: Includes specialized modules like an ExamLab for simulations, a homework planner, and a debate partner.