---
title: Overview
description: dlt supports both ETL and ELT transformation patterns
keywords: [elt, etl, transformer, transformations]
---

`dlt` supports both Extract, Transform, Load (ETL) and Extract, Load, Transform (ELT) patterns.

:::tip
Transform your data with `@dlt.hub.transformation` — eagerly on your compute or lazily as SQL pushed down to the warehouse, with incremental processing out of the box.
See the [dltHub transformations overview](dlthub.md) or follow the [full guide](../../hub/transformations/index.md).
:::

In ETL, the data is transformed before being loaded into the destination. This is useful for light processing such as adding columns, removing sensitive data, or type casting. `dlt` offers built-in utilities like `add_map()` and custom processors via `@dlt.transformer`

In ELT, the data is loaded as-is in the destination. This raw data is transformed directly on the destination where more powerful compute is available (e.g., data warehouse, data lake). `dlt` supports this via several patterns.

The two approaches can be used together in a single project.