BUD-E (Buddy) is an open-source voice assistant framework that facilitates seamless interaction with AI models and APIs, enabling the creation and integration of diverse skills for educational and research applications.

This repository is the implementation of the ProTACT architecture, introduced in the paper "Prompt- and Trait Relation-aware Cross-prompt Essay Trait Scoring" (ACL Findings 2023).

Educational language-learning app for Hokkien, a low-resource language, featuring flashcards, quizzes, and generative AI!

🤖 📝 An AI-powered essay writing assistant that helps you plan, research, and write well-structured essays using OpenAI and Tavily APIs. Built with Streamlit and LangGraph.

Three-layer memory architecture for long-term AI learning with Claude

MCA Method (Mentor, Copilot, Agent) - Pedagogical framework to transform LLMs into effective educational mentors for programming learning through structured guidelines.

An AI-powered notes summarizer and quiz maker with notebook-style summaries and voice support.

MEMOGEN: Master IELTS vocabulary with 3D interactive flashcards - 90%+ retention combining memory techniques, AI-generated imagery & LECTOR spaced repetition

An end-to-end, research-grade AI system for measuring human cognition. HCMS models mastery, confidence, learning stability, and adaptability through analysis, inference, validation, robustness testing, and explainability — bridging human-centered AI research and applied systems.

QuizGame TeamByteBrains - AI-powered terminal quiz game built with Jac language. Features multiple-choice questions with progressive hint system and Google Gemini integration for intelligent feedback after incorrect attempts.

Plataforma de aprendizaje colaborativo que utiliza un Teachable Agent (IA aprendiz) basado en RAG para fomentar el aprendizaje profundo y la analítica educativa en tiempo real.

Sorokin AI: Your Academic Mentor & Study Partner

All in One-Solution for converting documents to finetune LLMs

Modified DAMI hand-off model for LLM-based math coaching, trained on PRM800K and evaluated with GT-T metrics.