1. What ETL is and why it is needed

Operational systems store data for running the business; they are not designed for analysis:

• Data is spread across systems (ERP, CRM, Excel, APIs) with different formats and keys.
• Running heavy reports on OLTP databases slows down the business.
• History is overwritten in OLTP (customer moves city — old city lost).

ETL creates a single, clean, integrated, historical copy optimised for analysis.

[ERP SQL Server] ─┐
[CRM API]        ─┼─ Extract ─► Staging ─► Transform ─► Data Warehouse (star schema) ─► Power BI / SSRS
[Excel / CSV]    ─┘                                          ▲
                                               logs, audit, error tables