Use case

Lineage Mapping

Column-level and table-level lineage from source to consumer — the substrate every impact analysis, deprecation, and audit eventually needs.

Overview

Lineage is the answer to 'if I change this column, what breaks?' Without lineage that question takes a week. With lineage it takes a query.

What it solves

Enables impact analysis, deprecation, and audit at query speed instead of grep speed. Makes the cost of a schema change visible before the change ships.

How we build it

OpenLineage emitters on every transformation engine (dbt, Spark, Flink, Airflow) feed a lineage backend (Marquez, DataHub, Atlan, OpenMetadata). The graph is queryable; impact analysis runs against it; access boundaries from the source graph compose with lineage so 'who can see what derived from what' is a single question.

  • OpenLineage emitters on every engine
  • Table and column-level granularity
  • Impact-analysis queries on the graph
  • Integration with source-graph permissions

What changes when it is in place

Deprecations and migrations become safer. Audit becomes faster. The data team can give a confident answer to 'who is using this column' instead of a hopeful one.