CEDIS Learning Analytics: the whole ecosystem's data becomes pedagogical evidence — with privacy by design
The cross-cutting layer of the CEDIS Ecosystem gathers the signals from Slides, Atlas and Concept Studio into a pseudonymised, k-anonymous warehouse. Ingest and API are in production; the teacher dashboards are under construction.
Brasilia, August 9, 2026
Five products of the CEDIS Ecosystem record what happens in each class: readings, answers, questionnaires, concepts touched, routes travelled. The sixth is what makes the whole visible. CEDIS Learning Analytics, presented today as part of the CEDIS Ecosystem , works backstage: it gathers the signals of the whole ecosystem into a single analytical store, pseudonymised at the source, so that indicators, correlations and recommendations can come out of them.
It is the family’s cross-cutting layer — not the main tool of any teacher or student. It is born out of a 25-chapter discovery audit of everything the ecosystem already recorded, completed today, which fixed the architecture and a phased roadmap. The goal is the one on the infographic: engagement and progress indicators, detection of difficulties and advances, personalised recommendations and correlations between activities, concepts and performance — data that becomes better decisions.
Privacy comes before analysis. The student accepts, in the Learning Passport , a specific consent for analytics, and a dedicated analytical identity is derived in layers — it never coincides with the identity that Slides or the Atlas know. Revoking the consent propagates deletion all the way to the warehouse within a single ingest cycle. The API only returns aggregate views under K-anonymity of five or more; individual profiles are restricted to authorised services.
Plainly, so as not to sell what does not exist: ingest, the API and the deletion cascade are in production at analytics.cedis.tec.br. The teacher interface — the dashboards the infographic promises — is under construction. That is why the product enters the CEDIS catalogue as beta: the data foundation exists and works; the visualisation layer does not yet.
Learning Analytics is coordinated by Prof. Sergio Freitas and is open source under the MIT licence. The Learning Analytics page describes what is live and what is not; the big picture is in the CEDIS Ecosystem presentation .