How it works
Extract pulls data from applications and databases. Transform cleans, joins and reshapes it. Load writes it to a warehouse or lake for analytics. Modern stacks often use ELT, loading raw data first and transforming it inside the warehouse. ETL serves internal analytics; it usually runs in batches and is owned by data teams rather than product engineers.
Example
A finance team extracts invoices from the ERP and orders from Shopify every night, loads them into a warehouse, and joins them to build a revenue dashboard.
In integrations
ETL tools move data into your warehouse for your own analysis. Unify Int solves a different problem: customer-facing integrations inside your product, with auth for each of your customers, write-back and webhooks. The two can coexist, and normalized data from the Unify Graph is easier to load downstream. If your goal is internal reporting, ETL is the right tool. If your goal is a feature your customers switch on in your product, you need customer-facing integrations instead.