What is ELT?
ELT loads raw data into the warehouse first and transforms it there afterwards, using the warehouse's own processing power.
ELT explained
Because cloud warehouses can process large volumes cheaply, many teams copy source data in untouched and then build clean tables from it with SQL. The raw copy stays available, so a transformation can be corrected and re-run without going back to the source.
The trade-off is that raw, possibly sensitive data lands in the warehouse before it is cleaned, so access to the raw layer needs controlling.
Example
A pipeline copies all rows from the billing database into a 'raw' schema every hour; scheduled SQL then builds a tidy 'invoices' table for analysts.
How it differs from similar terms
In ETL the transformation happens before loading.