Skip to content
BENIBIX Learn

Modern Data Stack

Data🧰 PlatformsData Platforms

Also known as MDS, Data Stack, Cloud Data Stack

A data platform built from cloud tools, each doing one job.

Think of it like

It is like a band. Each one plays a part, and all of them play one song.

Example

A startup loads data with Fivetran and keeps it in Snowflake. It cleans it with dbt and shows it in Looker.

Why it matters

A small team can stand up a full platform in days. Each part can be swapped later.

Definition

The modern data stack is a common way to build a data platform from separate cloud tools, each doing one job. An ingestion tool copies data in. A cloud warehouse stores it. A transform tool, often dbt, cleans it with SQL. A BI tool shows it, and an orchestrator runs the steps in order. The name took off around 2020, as these tools grew popular.
The modern data stack is this row of boxes, each one a separate tool.

Common mistake

Adding a new tool for every small need. Soon there are fifteen bills and nobody owns how they fit.

Key takeaways

  1. 1
    One job each. Load, store, change, show.
  2. 2
    In the cloud. Bought as services, not servers.
  3. 3
    Swappable. Change one part, keep the rest.