# ETL Overview

Extract Transform Load (ETL) generally involves the transferral of data from one (or more) sources to another, including a certain amount of 'transformation' of the data so that it adheres to the required protocols of the destination system

**Extract Transform Load (ETL)** generally involves the transferral of data from one (or more) sources to another, including a certain amount of 'transformation' of the data so that it adheres to the required protocols of the destination system.
Some different types of ETL implementation are:

* Extract data from a source, run a **simple transformation (e.g. set data type) in Tray**, then load to a database of your choice
* Extract data from a source, store it in an intermediary (such as AWS EMR) which can act as a **staging environment which can perform more 'heavyweight' transformations**, then load to a database of your choice
* Extract data from a source, load straight to a database of your choice, **then perform transformations directly in the new database environment** (this is more of an ELT implementation)
  Some **common SaaS sources** used in ELT are:
* CRM systems e.g. [Salesforce](https://tray.ai/documentation/connectors/service/salesforce/)
* Automated marketing systems e.g. [Marketo](https://tray.ai/documentation/connectors/service/marketo/), [Outreach](https://tray.ai/documentation/connectors/service/outreach/)
* HR systems (employee and payroll info etc.) e.g. [BambooHR](https://tray.ai/documentation/connectors/service/bamboohr/)
* Project management systems e.g. [Asana](https://tray.ai/documentation/connectors/service/asana/), [Workfront](https://tray.ai/documentation/connectors/service/workfront/)
  Most commonly the data is finally **loaded into databases / data warehouse solutions**, i.e.:
* MySQL, PostgreSQL, MSSQL
* BigQuery, Snowflake, Redshift