Analytics Engineering

Purpose

Design and build production analytics pipelines with dbt Core (models, materializations, Jinja macros, ref/source, tests, documentation), metrics and semantic layers (dbt Metrics, MetricFlow, Cube.js, metric definitions, dimensions, filters, time granularity), data modeling for analytics (marts approach, One Big Table, dimensional modeling, medallion architecture), and analytical SQL (window functions, CTEs, pivoting, statistical functions, time series, performance optimization, UDFs).

Agent Protocol

Trigger

Exact user phrases: "analytics engineering", "dbt", "dbt model", "dbt materialization", "Jinja macro", "dbt ref", "dbt source", "dbt test", "dbt doc", "metrics layer", "semantic layer", "MetricFlow", "Cube.js", "dbt metrics", "data modeling", "marts approach", "OBT", "One Big Table", "dimensional modeling", "medallion architecture", "bronze silver gold", "SQL analytics", "window function", "CTE", "pivot", "unpivot", "analytical SQL", "time series SQL", "SQL UDF".

Input Context

Before activating, verify:

• Transformation tool (dbt Core, dbt Cloud, SQLMesh)
• Data warehouse (Snowflake, BigQuery, Redshift, Databricks, Postgres)
• BI tools (Tableau, Looker, Power BI, Metabase)
• Existing data model layer (raw, staging, intermediate, marts)
• dbt version and packages installed (dbt_utils, dbt_expectations)
• CI/CD setup (GitHub Actions, dbt Cloud CI)
• Testing and documentation practices