A university's data architecture is usually a portrait of its org chart. Each function acquired the system it needed. Each system is competent within its boundary. Almost none were designed to answer a question that crosses boundaries — which is where nearly every important institutional question actually lives.
01The join that nobody owns
Asking why students in a particular program persist at higher rates requires curriculum sequencing, engagement patterns, advising history, financial context, and outcome data. Those live in five systems and three departments. The analysis is possible, but it is a project — commissioned, scoped, staffed, and delivered months later, after which the conditions have changed.
Your institution is constantly creating data. An intelligence layer is what lets it learn from itself.
02What an intelligence layer is not
- 01Not a replacement for the SIS, LMS, or data warehouse. It reads from them.
- 02Not another interface for staff to check. Intelligence should arrive where decisions are already being made.
- 03Not a governance shortcut. It raises the stakes on access, consent, and purpose limitation, and should be designed accordingly.
Artifact Research · Artifact Intelligence
Artifact Research publishes the working thinking behind the Learning Intelligence Platform, including the parts that are still open questions.