fabricks-glossary

Company Glossary — Data Warehouse (Databricks)

This skill provides domain vocabulary, entity definitions, and architecture context for the company's data warehouse, built on Databricks using the Fabricks.Runtime framework. Most of this terminology originates from and is primarily used in the context of that data warehouse.

How to use this skill

Load only the reference file(s) relevant to your current task:

Topic	Reference file	Load when...
Acronyms & abbreviations	references/acronyms.md	An unknown abbreviation appears (SPT, BU, IVP, HEP, etc.)
Core business entities	references/entities.md	Working with articles, customers, sales agents, or inventory
Domain concepts	references/domain-concepts.md	Working with commissions, margin bonuses, campaigns, credit risk, or MDM
Naming conventions	references/conventions.md	Naming or interpreting tables, columns, schemas, or data layers
Fabricks framework	references/fabricks-framework.md	Working with steps, jobs, modes, CDC, or pipeline configuration

Quick reference — most-used facts

• Table prefixes: dim_ dimensions · fact_ measures · ref_ lookups · scd1_/scd2_ history · merger_ consolidated · *_agg aggregated · *_hist snapshots
• Key column suffixes: __key hash key · __identity surrogate · __is_current SCD2 flag · __valid_from/to SCD2 range · _sk foreign surrogate · _pct percentage
• Data layers: staging (raw ingest) → raw (SCD applied) → transf/core/mdm/ml (gold) → semantic/bme (presentation) → distribution (export)
• Core schemas: core has surrogate keys (__identity); transf has business logic but no core tables; semantic uses UPPERCASE column names with spaces for Power BI