Ed Vs Spark
The North Star of Spark
Spark's single driving goal is to optimise for smaller scales without compromising DPI architectural principles, or the ability to scale on demand. Spark will still scale for larger user bases if needed. Everything in Spark is in service of cost reduction, simplification, and modernisation.
WHAT STAYS THE SAME
Core Building Blocks (Retained as in SB Ed - not replaced)
Sunbird Knowlg
✅ Retained — merged but not replaced
Sunbird Lern
✅ Retained — merged but not replaced
Sunbird inQuiry
✅ Retained — merged into Knowlg service
Sunbird RC v1.x
✅ Explicitly kept as-is for Verifiable Credentials
Keycloak
✅ Retained for identity & auth (now with Yugabyte backend and upgraded to OIDC protocol enabling replacing keycloak with any other tool)
Kong API Gateway
✅ Retained (upgraded, not replaced)
Kafka
✅ Retained (upgraded to v4.x)
Apache Flink
✅ Retained — explicitly confirmed "No" to removing Flink
Content Players
✅ All existing players reused (PDF, Video, H5P, ECML, EPUB)
Collection & ECML Editors
✅ All editors reused with existing workflows
Telemetry Service
✅ Retained as part of default install (even when Obsrv is add-on)
QuML/inQuiry Player & Editor
✅ Retained
Certificate issuance and validation flow
✅ Retained (Sunbird RC v1)
Multi-tenancy via Channels
✅ Same logical multi-tenancy model
Offline consumption (mobile)
✅ Retained, now explicitly in scope for Spark
Completely new Android consumption app and portal
Core Functional Capabilities Retained
Course creation, batch management, enrollment
Certificate issuance & passbook
Content creation workspace
Content review workflows, including async publish workflow
User profile & learning records
Global search
Anonymous content consumption
RBAC & user management
Notifications
I18N / multilingual support
Instrumentation & telemetry generation
Last updated
Was this helpful?
