Member-only story
How to transform Legacy SAS ETLs to DBT
Understanding dbt’s Role
dbt is primarily designed for SQL-based transformation workflows, and it works best with modern data warehouses like Snowflake, Redshift, BigQuery, and Databricks. It does not natively support SAS as a backend.
Potential Integration Approaches
Since SAS is more than just SQL (it includes data step processing, macros, and procedural language), the team would need to adapt their workflow:
a. Using dbt with SAS via SQL-Pass Through
• If your SAS environment is connected to a relational database (e.g., Teradata, Oracle, SQL Server, or Netezza), you can write dbt models in SQL and use SAS/ACCESS to execute them.
• The dbt models would be designed to transform data in a database, while SAS would still be used for reporting or additional processing.
b. Exporting Data to a Modern Data Warehouse
• If your team is moving away from SAS for ETL but still needs SAS for analytics, dbt can be used to transform the data in Redshift, Snowflake, or BigQuery before exporting it to SAS.
• You can use Python, ODBC, or CSV-based exports to move data between dbt and SAS.
