AieDu
B2B · B2G · B2C Coming Soon

Learn directly from the AI that coded it
AI education.

The AI that codes a software project understands it better than anyone. We taught that coding AI pedagogy — and built a curriculum where it teaches you directly. In the age of vibe coding, learn from an AI expert.

Contact UsTry Demo

Government & Public Institutions (B2G)

Enterprise AI Training (B2B)

Educational Institutions (B2B)

Individual & Freelance Educators

Students & Parents [B2C Coming Soon]

AI Research Institutes

Trusted By

AIRI

How Is It Built?

Three contexts are synthesized so AI can automatically generate the curriculum.

Source Context

Analyze the Subject

AI directly analyzes the source — SW projects, textbooks, educational programs, academic papers. Domain doesn't matter.

Pedagogy Context

Select the Teaching Method

Choose and combine from a pedagogy library — Socratic method, ITS hint ladders, and more — matched to the goal.

Platform Context

Reflect the Environment

Practice panel configuration (code editor, math graph, science simulation) and deployment mode are embedded into the curriculum.

AI auto-generates → man-in-the-loop review → learning begins

How Does the AI Teacher Teach?

Not just an AI that answers. Pedagogy benchmarked from CS50P, Google ML, and Khanmigo — then redesigned for the AI age.

Socratic DialogueITS 4-Level Hint LadderDirect Teaching (Panel Control)Adaptive Pacing
Socratic dialogue

No direct answers

When code is wrong, the AI doesn't hand over the fix. It starts with "What do you think this code will do?" and leads the student to find the issue themselves. Modeled after Khanmigo and CS50.ai's AI tutor design — real dialogue, not Q&A.

ITS research-based

4-level hint ladder

Stuck? Direction hint → concept hint → analogy → minimal explanation — in that order only, no skipping. Based on Intelligent Tutoring System (ITS) research showing that early answers reduce learning retention.

Redesigned for the AI era

You're learning AI, not just Python

Built on CS50P's progressive scaffolding and Google ML Crash Course's applied examples. Every concept has an AI anchor: learn loops and you understand training epochs; learn functions and you understand AI pipelines.

Pedagogy is not fixed. Methods are selected and combined from a library to match the objective.

Design references:Harvard CS50PGoogle ML Crash CourseKhanmigoCS50.aiITS Research

Platform Features

Core

AI Teaches Directly

Not a Q&A bot. The AI teacher reads the curriculum, manages progress, and adapts to each student in real time.

Core

Auto-Generated Curriculum

Change the source and a new curriculum is produced — no manual authoring required. When the source changes, the syllabus updates instantly.

All customers

Curriculum Plugin

From Python basics to Naia OS, OpenClaw, JiKiME ADK — your institution's curriculum can plug right in.

B2B·B2G

On-Premises Deployment

Air-gapped environments supported. Deploy on your own GPU servers for fully offline, self-contained operation.

B2B·B2G

Student Activity Logging

Code executions, AI questions, and progress are all logged. Measure educational outcomes with data-backed reports.

B2B·B2G

Per-Institution Customization

Customize the AI teacher's name, tone, and safety policy per client. Air-gapped deployments automatically enforce strict safety mode.

Roadmap

AI Context Export

Learning outcomes are exported as a structured context. External AI assistants (e.g., Naia OS) receive this context and continue supporting students in real work.

Roadmap

Omni Model AI Teacher

Integrated with Naia OS Omni model for multimodal teaching — voice, video, and immersive learning with a VRM avatar.

Considering a deployment?

Questions about curriculum integration, on-premises setup, or licensing? Reach out via email.