µEd: A Shared API for Education Microservices — Accepted at L@S '26

The work-in-progress paper µEd API: Towards a Shared API for Education Microservices has been accepted at the 13th ACM Conference on Learning @ Scale (L@S 26) and will be presented in Seoul, South Korea, as part of the joint Festival of Learning 2026.

Education platforms today often lock institutions into a single ecosystem. Whoever wants AI-powered feedback, assessment, or chat-based tutoring is left with whatever the platform happens to offer — or has to build a custom integration for every new tool. Learning at scale depends on exactly this kind of domain-specific automation, so the choice of platform quietly turns into a choice about what is pedagogically possible.

The paper proposes µEd, a shared, standardized interface that lets specialized microservices plug into any learning platform. The initial specification covers three service types — automated feedback, assessment, and educational chatbots — with further additions, such as learning analytics, planned for the future.

µEd was built collaboratively across four institutions — the Technical University of Munich, Imperial College London, ETH Zürich, and Nanyang Technological University, Singapore — by comparing the systems each of them already runs in production and abstracting out the common core. All four are now adopting the API, which grounds the specification in real deployments rather than in a greenfield design.

As a work-in-progress contribution, the paper is an invitation: teams building or running educational technology are welcome to try the specification, give feedback, and help shape where it goes next. The API and its ongoing development can be followed at mued.org.

Citation