ELT (Extract, Load, Transform) is the modern variant of classic ETL. Data is extracted from the sources and loaded directly into the data warehouse in its raw format; the transformations are applied at the destination, typically in SQL on the data warehouse engine (Snowflake, BigQuery, Redshift). The order of the letters makes all the architectural difference.
ELT gained ground because cloud-native data warehouses have unlimited compute power on demand. Transforming at the destination lets you keep the raw data (useful for later re-analysis with new questions), parallelise heavy transformations, and use SQL tools familiar to analysts instead of specialised ETL software. The ELT trend has been dominant in new implementations since 2020.
For INFOS customers with larger data volumes or more complex analytical requirements, the combination of Qlik Sense + a cloud data warehouse (via an ELT pipeline) is the reference architecture. For smaller operations, traditional ETL remains valid — less complexity, less infrastructure cost, sufficient performance for the needs.