OLAP is the acronym for Online Analytical Processing. It is a set of techniques for the rapid and multidimensional analysis of big data, starting with the organization of specific databases.

The OLAP methodology is related to business intelligence, as it implies the company ability to obtain useful information from the data available with specific tools. In particular, OLAP was created to speed up the process of reading, analyzing and recovering data through a different database structure and organization.

In standard databases, data is stored in spreadsheets and two-dimensional tables, not suitable for multi-dimensional analyzes. OLAP databases use cubes instead of tables. These cubes are calculation structures in three or more dimensions instead of 2D.

The OLAP process extracts data from multiple sources, heterogeneous and unrelated to each other, stores them in simple data warehouses but then cleans them, organizes them in semantic models and permanently stores them in cubes.
In each cube, the measures, that is the numerical data, are organized by dimensions.

These dimensions are categories that can be, for example, geographical or temporal. These dimensions-categories are structured by level hierarchies: tree models, in which data and groups of data, called members, are classified into relationships of similarity, familiarity and derivation, the so-called "parent-child" relationships.

Regarding great OLAP tools, Apache Kylin is a multidimensional open-source analysis engine. It is designed to provide SQL and MOLAP interface in sync with Hadoop to support large datasets. It supports quick three-step query processing; identify the star pattern, create a cube from the data tables, and query and get results via API. Kylin was developed to reduce query processing time for faster processing of billions of rows of data.

Then, Swiss-based company icCube owns a business intelligence software with its name. It sells an online analytics processing server implemented in Java according to J2EE standards. It is an in-memory OLAP server and is compatible to work with any data source that stores its data in tabular form.

IcCube comes with built-in plug-ins that facilitate file access and HTTP streaming, etc. It has a unique web interface to perform tasks such as cube modeling, MDX (multidimensional expression) queries, server monitoring and dashboards. It is an excellent data analysis and visualization tool with a focus on quality.

Then, Pentaho is a powerful open-source tool that provides key BI capabilities such as OLAP services, data integration, data mining, extract-transfer-load (ETL), reporting, and dashboard functionality. Pentaho is built on the Java platform that can work with Windows, Linux and Mac operating systems.

Pentaho is available in two editions: one is Enterprise Edition and another is Community Edition. The Enterprise edition has extra support features and services. It is a highly flexible BI tool with awesome comprehensive capabilities.

Last but not least, Mondrian is a very interactive tool with exceptional features and strengths such as its ability to work with categorical data, large data and geographic data. It is a general-purpose data visualization tool. It consists of interconnected charts and queries.

Mondrian initially focused primarily on categorical data visualization techniques. However, over time, a full suite of visualizations for univariate and multivariate data has been added. Its link to R offers excellent statistical procedures.