In A Nutshell

Pasted image 20260726222408.png

Data Warehouse

Data warehouse are the central data hub for the analytical and reporting needs of an organization.
The data in the a data warehouse is structured, and predefined schema (enforced during the ETL/ELT pipeline)

Bill Inmon originated the notion of the data warehouse, which he described as "a subject-oriented, integrated, nonvolatile and time variant collection of data in support of management's decisions."

Data Mart

Data marts are a subset of a Data Warehouse, where data warehouse hold the data across the organization, the data mart are specific to a department or function. It could be specific to Sales, CRM, etc.
Data Mart exists for two reason -

  1. It make the access of data easy for the Data Analyst and Reporting folks.
  2. It can also provide an additional transformation, beyond the initial ETL/ELT pipelines, that can significantly improve performance, if reports or analytical queries requires complex joins and aggregations of data.

Operational Data Store (ODS)

ODS is a type of data warehouses, designed for near realtime reporting and operation decision making, positioned between data source (OLTP) and the data warehouse They are mainly used when the analytics are required within minutes of timeframe.
They are optimized for queries that reflects the present state of the business, rather than long term trends.

Data Lakes

Data Lake store the data in it's raw / original (as-is) form without any pre defined schema, accepting structured, semi-structured and unstructured data. That is the data sources can be relational database, IoT devices, hierarchical data.
Data Lake applies schema-on-read. This flexibility make the data lakes well suited for storing large volumes of raw data for machine learning and data science exploration.

The trade of is reliability, without any enforcement of pre defined schema, the data could be inconsistent, duplicate, or poorly documented data that can not be trusted for analysis or governed reporting.

Lakehouse

Lakehouse or Data Lakehouse is a system of data management which is combined by the flexible, low cost storage of a data lake and reliability, governance and query performance of a data warehouse, that supports both BI and ML/Data Science.

It achieve that by

  1. metadata layer (like Delta Lake) for the data lake

References