Select a technical architecture
Before designing a data model, select a technical architecture. This topic uses MaxCompute together with DataWorks to walk through the complete data modeling and development workflow.
The following diagram shows the full architecture.
Architecture components
The architecture has two core components:
MaxCompute is the offline compute engine for the entire big data development process. It handles large-scale data processing and analytics.
DataWorks manages data movement and pipeline orchestration through the following integrated functions:
-
Data Integration — collects data from source systems and performs basic extract, transform, and load (ETL) operations
-
Data Development — builds and schedules data pipelines
-
Data Quality — monitors and validates data at each stage
-
data security — protects data assets across the platform
-
Data Management — governs data access, lineage, and compliance across the platform
Data flow
Data moves through the platform in three stages:
-
Ingestion — Data Integration pulls raw data from source systems into MaxCompute.
-
Processing — Data Development transforms and refines raw data into clean, business-ready datasets.
-
Serving — Processed data flows to analytics, reporting, and downstream applications.
What's next
After selecting a technical architecture, create a data model using MaxCompute and DataWorks.