University × AI Infrastructure

The learning layer for university AI.

We’re building a connection between your university LMS and the AI tools you use—so course materials stay available without downloading and uploading them for every conversation.

Explore the architecture

In development · Starting with Chung-Ang University e-Class on macOS

01 — THE EVERYDAY GAP

Your next conversation
should start with your course.

Today, you carry the context.

Find the material in your LMS. Download it. Upload it to AI. Explain which course it belongs to. Repeat when something changes.

We’re building a connection that carries it.

Connect your university account. Keep course materials synchronized. Let a connected AI tool retrieve the context when you need it.

02 — SYSTEM

From LMS
to AI context.

The planned system connects your university account locally, synchronizes course materials to UnivMCP Cloud, and makes them available to connected AI applications. Your university credentials stay on your device.

01
University Systems
Your university LMS remains the source of your courses and materials. The first release focuses on course resources, starting with Chung-Ang University e-Class.
02
Local Collector
The collector is designed to connect from your device and collect the course resources your account can access. University credentials stay local.
03
Synchronized Academic State
Original course files and their metadata are kept together in the planned cloud layer, so you can return to your course context across AI conversations.
04
MCP / API → AI
Model Context Protocol (MCP) and API interfaces are the planned access point for compatible AI tools to find your courses and retrieve your synchronized materials.

Read the system design

03 — DIRECTION

Your courses.

Context that carries forward.

A foundation for university AI.

04 — WHAT THIS ENABLES

One connection.
A new class of university software.

When academic systems become continuously accessible to AI, the opportunity extends far beyond file retrieval. Course-aware assistants, adaptive learning support, cross-course reasoning, academic automation, learning analytics, and entirely new interfaces become possible.

UnivMCP is being built as the connective layer beneath them.

05 — STATUS

We are building it now.

STATUS / EARLY DEVELOPMENT
SPECIFICATION / IN PROGRESS
REFERENCE IMPLEMENTATION / IN PROGRESS
UNIVMCP.COM / 2026

Built from everyday use.

The project grew from a CAU e-Class MCP tool the maintainer uses with Codex. That early prototype informs this work; UnivMCP is developing a broader local collector, cloud sync, and remote retrieval architecture.