Business Intelligence (BI)

Dashboards and reports that help people make decisions.

Data Engineer · Junior Data Analyst · Junior · core

Also known as: BI, analytics reporting, business analytics

Business intelligence (BI) is the practice and tooling for turning company data into reports and dashboards that help people make decisions. Typical questions: what were sales last quarter, which regions are growing, how many users churned.

A typical BI stack, from the data side:

source systems ─▶ pipelines ─▶ warehouse ─▶ BI tool (dashboards, reports)

BI tools include Looker, Tableau, Power BI and Metabase. They connect to the warehouse, run queries, and present charts that non-programmers can filter and explore.

BI versus other analytics work

BI mostly looks backwards at known questions with agreed metrics (“what happened?”). Data science and machine learning explore open-ended questions or predict the future. Both depend on the same clean data underneath.

What data engineers do for it

Prepare modelled, documented tables designed for analysis.
Define metrics once so charts agree. See metric definitions and the semantic layer.
Keep data fresh and monitor the pipelines.
Control cost: poorly written dashboards can run heavy queries constantly.
Enable self-service so analysts can answer their own questions without a ticket to the data team.

Pitfalls

Conflicting numbers. The classic BI problem: finance and sales report different revenue because they define it differently. Settle definitions and implement them in one place.
Trusting a number without knowing its source. Show where the data comes from and when it last updated.
Too many dashboards, too little adoption. Build for real decisions, check usage, and retire what nobody opens.
Treating the BI tool as the data model. Putting logic only inside dashboards makes it hard to reuse and test.