HKUDS/DeepTutor

DeepTutor: Lifelong Personalized Tutoring. https://deeptutor.info/.

What it solves

DeepTutor addresses the need for a personalized, lifelong learning assistant that can manage complex educational workflows. It helps users organize knowledge, automate research, and interact with AI-driven "Partners" that retain context and skills over time, moving beyond simple chat interfaces to structured, goal-oriented tutoring.

How it works

DeepTutor functions as an agentic framework that connects various AI models and tools into a unified workspace. It uses advanced retrieval techniques (such as LightRAG and GraphRAG) to process documents, books, and web sources into a searchable knowledge base. Users can create custom "Partners"—specialized AI agents with their own memory and skills—to assist with specific subjects, research tasks, or guided learning paths. The system handles document parsing, multi-format file support, and secure, isolated runtimes for executing code or tools.

Who it’s for

It is designed for students, researchers, and lifelong learners who want to build a private, self-hosted AI tutoring environment. It is also suitable for developers or power users who want to integrate their own knowledge bases, GitHub repositories, or third-party AI providers into a cohesive learning dashboard.

Highlights

  • Agentic Learning Partners: Create and assign specialized AI agents that maintain private memory and unique skill sets.
  • Advanced Knowledge Management: Supports diverse data sources including PDFs, EPUBs, GitHub repos, and web content, indexed via modern RAG techniques.
  • Guided Mastery Paths: Features structured learning dashboards that track progress and gate content based on readiness.
  • Extensible Ecosystem: Includes a CLI, plugin support, and the ability to connect various LLM providers and external tools.
  • Privacy-Focused: Offers self-hosted deployment options with user-level isolation and secure workspace management.

Written about in

Related

  • Project
    xiao18825501901-rgb/coursemate-ai

    A personalized teaching platform that combines RAG-based course material retrieval with an AI agent for task management and study planning.

  • Project
    Alibaba-NLP/VRAG

    A multi-turn multimodal RAG framework that uses agentic reinforcement learning to enable VLMs to reason and retrieve information across text, images, and videos.

  • Project
    amosblomqvist/learnharness

    LearnHarness is an open-source agent harness that teaches you through AI agents: it measures what you understand, plans a lesson step by step, and builds a persistent knowledge graph of your reasoning so learning carries across sessions.

  • Project
    GiovanniPasq/agentic-rag-for-dummies

    A tutorial‑style, open‑source Python project that shows how to build a modular, agent‑driven Retrieval‑Augmented Generation system with LangGraph, hierarchical chunking, conversation memory, query clarification, multi‑agent map‑reduce, self‑correction, and evaluation via RAGAS. Works with Ollama, OpenAI, Anthropic, Google Gemini, Qdrant vector store, and optional Langfuse observability.

  • Project
    madhvantyagi/Gnos

    A teaching harness that turns coding agents into personalized tutors by creating adaptive, multi-modal courses based on user goals and adaptive learner tracking.