How sunbird-knowlg/knowledge-platform works
Project type and purpose
• This is a microservice-based backend platform (“Knowledge Platform 2.0”) exposing domain APIs (Content, Assessment, Search, Taxonomy, etc.) via Play/Netty servers.
• It uses Akka actors for business logic, Cassandra/Neo4j/Redis/Elasticsearch for data/backing stores, Kafka for messaging, and shared libraries for common concerns.
File structure
• Top-level per-domain folders (content-api, assessment-api, search-api, taxonomy-api, ontology-engine, platform-core, platform-modules)
• Each domain splits into an “actors” module (the Akka actor layer) and a “service” module (HTTP controllers, filters, DI modules)
• Shared libraries in platform-core (actor-core, cassandra-connector, kafka-client, cache, common utilities, telemetry, schema-validator)
• platform-modules for off-line/content import, MIME‐type and URL management
• Infrastructure-as-code under build/ (Dockerfiles, CI pipelines), docker-compose.yml, kubernetes/helm_charts/, terraform/
Key insights from README
• All services are run locally via Docker (Cassandra, Neo4j, Redis, Kafka, Elasticsearch)
• Each service starts with “mvn play2:run” on port 9000 and exposes a /health endpoint
• Local setup script automates container provisioning