For IT, instructional design, and faculty leads
Technical overview
What we expose on purpose: how courses ship, where code runs, and how learning events become data you own.
Course Pack
Authoring pipeline produces portable IMSCC for LMS import; materials stay versioned so upgrades don’t fork silently.
- Jupyter Book → export → IMSCC
- LMS import validation checklist (your team + ours)
Lab Course
GitHub org per institution; repo-per-assignment; Codespaces definitions checked into the same repo students fork.
- Codespaces configs (pinned images, secrets policy)
- CI for autograding contracts
AI Faculty
Runtime services back Dialogic, SAMWISE, and BEATRICE—scoped to course rosters and policy you set.
- Dialogic — lecture orchestration & interactivity
- SAMWISE — tutor endpoint (per learner)
- BEATRICE — grading agents + feedback pipelines
Insight Layer
xAPI-native: statements describe interactions; you attach an LRS (local or cloud) that your security team approves.
Endpoint (illustrative): xapi.castalia.institute — exact paths and keys provided under pilot agreement.
What we don’t claim
We’re not “an LMS.” We’re AI Faculty infrastructure: deploy courses, upgrade autonomy, and measure outcomes with open standards—so procurement can compare us to courseware without locking learning data in a black box.