EduOrchestrate is a multi-agent AI platform designed to orchestrate personalized learning workflows for students and academic institutions. Built with LangGraph and LLMs, it coordinates specialized AI agents — tutor, planner, evaluator, and resource finder — into a seamless, adaptive education pipeline.
EduOrchestrate aims to solve the personalization gap in digital education by deploying a graph-based orchestration layer that coordinates multiple AI agents, each specialized in a different aspect of learning — from concept explanation and quiz generation to study schedule planning and resource curation. The platform is being developed with IBM Watsonx Orchestrate and LangGraph for enterprise-grade agent coordination.
Explains concepts at the student's level, answers follow-up questions, and adapts explanations based on learning history.
Generates personalized study schedules based on syllabus coverage, upcoming exams, and daily availability constraints.
Creates topic-specific quizzes, grades responses, tracks mastery scores, and flags weak areas for targeted revision.
Curates papers, videos, and references matching the student's current topic, proficiency level, and learning style.
Coordinates agent interactions using a state graph — routing tasks, managing memory, and resolving inter-agent dependencies.
Maintains long-term student profiles, session histories, and contextual state across all agents for coherent multi-turn interactions.
Student provides their course syllabus, learning goals, available hours, and preferred learning style to initialize their profile.
The orchestrator builds a state graph with nodes for each agent and defines conditional edges based on the student's context and intent.
Each student request is classified and routed to the appropriate agent (or chain of agents) based on intent detection and current state.
Agents collaborate — e.g., the Tutor explains a concept, the Evaluator immediately generates a micro-quiz, and the Planner updates the schedule.
Session results update the student's long-term memory store, refining future agent responses and improving personalization over time.
LangGraph state machines coordinate agents with conditional branching, loops, and shared memory — enabling complex workflows.
Each agent is independently developed and tested, enabling easy replacement or extension without breaking the overall pipeline.
The system adapts to each student's pace, strengths, and weaknesses — delivering a unique learning path every session.