# Analytics engineering · Glossary
The discipline of transforming raw data into clean, tested, documented models that analysts and business users can rely on.
[Glossary](/glossary) · Analytics engineering and analytics

# Analytics engineering

     The discipline of transforming raw data into clean, tested, documented models that analysts and business users can rely on. It applies software engineering practices, such as version control, testing and code review, to analytical data.

## Why it matters

     Analytics engineering sits between data engineering and analysis. It is where business definitions, such as revenue, active customer or on-time delivery, are encoded once and reused, rather than recalculated differently in every report.

     Treating transformations as tested code makes numbers reproducible and changes reviewable, which builds trust in reporting and provides the stable foundation that AI and self-service analytics need.

## In practice

     For example, a UK hospitality group might define covers, average spend and labour cost ratio once in version-controlled models with automated tests, so every venue report, forecast and board summary uses identical logic.

## Where Rodan fits

     Rodan’s [Data and Analytics Engineering](https://rodan.io/data-analytics-engineering) discipline builds tested analytical models. See also [what building Quantsole taught us about production analytics](https://rodan.io/insights/what-building-quantsole-taught-us-about-production-analytics).

## Related terms

- [Data modelling](/glossary/data-modelling)

- [Semantic layer](/glossary#semantic-layer)

- [ETL (extract, transform, load)](/glossary/etl)

- [Data engineering](/glossary/data-engineering)

- [Single source of truth](/glossary/single-source-of-truth)

- [Dimensional modelling](/glossary/dimensional-modelling)

- [Medallion architecture](/glossary/medallion-architecture)
HTML: https://rodan.io/glossary/analytics-engineering