Products

Not a set of tools,
but a learning system.

Overview

COCOROBO builds AI-native learning systems for real classrooms. The goal is not to generate content, but to structure learning processes.

System Architecture

COCOROBO operates as a layered learning system. CocoFlow provides the AI infrastructure layer. CocoClass, CocoNote, and CocoStudy form the application layer. CocoPi extends into the physical world. All components are interconnected.

CocoFlowAI InfrastructureCCocoClassClassroomNCocoNoteCollaborationSCocoStudyPersonalPCocoPiPhysicalIn-class ↔ Out-of-classDigital ↔ PhysicalHuman ↔ AI

CocoClass

In-class orchestration

Teaching → learning → feedback → adaptation.

CocoNote

Collaboration & knowledge building

Shared artifacts, group cognition, knowledge building.

CocoStudy

Personalized learning

Out-of-class adaptive feedback and development tracking.

CocoFlow

AI infrastructure

Multi-agent construction, workflow orchestration, RAG, HITL.

CocoPi

Physical extension

AI + sensors + IoT, project-based learning.

The CocoClass × CocoStudy Learning Loop

CocoClass handles in-class orchestration: AI understands the teaching intent of each lesson page, automatically assembles interactive elements, and provides real-time classroom analytics — a Teacher Control Room that gives educators a panoramic view without changing their workflow.

CocoStudy handles out-of-class learning: starting from AI-powered error diagnosis, it generates personalized learning paths with Socratic questioning, scaffolded practice, and variant exercises — transforming from a "blind practice system" into a personalized diagnostic learning system.

Together they form a continuous loop: classroom performance determines today's homework; homework results guide tomorrow's lesson.

TTeachingLLearningFFeedbackAAdaptation

The SMART Principles for Future Classrooms

Based on a decade of practice and grounded in learning sciences, we propose the SMART framework — a set of design principles for future classrooms. This is not a feature checklist, but a system-level implementation of learning science. Most AI education tools can only "generate materials" but cannot "manage the learning process." SMART addresses this gap.

SMART

S— Sharing

Making thinking visible — ideas, processes, and reasoning are externalized and shared across the classroom.

M— Measurement

Capturing learning processes through data — enabling teachers to see, interpret, and respond in real time.

A— Adaptation

Responding to individual and group differences — adjusting instruction, content, and pace dynamically.

R— Reconstruction

Deepening understanding through iterative cycles — revisiting, refining, and rebuilding knowledge.

T— Teamwork

Learning as a social process — collaborative construction of knowledge through structured group interaction.

SMART is embedded throughout our system — from CocoClass (Sharing), to CocoStudy (Measurement), to CocoNote (Teamwork), to CocoFlow (Adaptation), and the whole system's iterative Reconstruction.

Vision

The next stage is AI-native learning systems — where learning becomes continuous, observable, adaptive, and collaborative.

Next: Thinking