An example transformation requirement might be to take all of the incoming customer orders, clean up the data for consistency, and aggregate it into a

How DBT DevOps Enables Data Teams

submited by
Style Pass
2021-09-06 10:00:06

An example transformation requirement might be to take all of the incoming customer orders, clean up the data for consistency, and aggregate it into a “sales by region” summary table for our business users.

A further example might be to then take this aggregated sales data, identify products low in stock in each region, and produce data in the correct format to place an order with our supplier.

Companies have been doing this type of Extract, Transform and Load (ETL) work for years, and it’s a very mature field with established software and patterns. However, DBT’s approach is really a huge step forward in modernising the process for the cloud database age.

Historically, businesses used a process referred to as ETL to populate data warehouses. This involved Extracting data from source applications and databases, Transforming it into the required formats, and then Loading it into the warehouse for consumption.

DBT flips this process around by executing the transformations directly within the database or data warehouse, after it’s been loaded. The process therefore changes to Extract, Load and Transform or the acronym ELT.

Leave a Comment