Quickstart

Choose a source

Open Settings → Integrations, select a warehouse, database, cloud catalog, or pipeline source, and choose the projects, catalogs, datasets, or schemas Embrasure should index.

Use the supported sources matrix to choose the evidence path for your stack.

Grant read-only access

Create a dedicated identity with access only to the selected scope. Prefer managed identities, workload identity, OAuth, or key pairs over shared passwords.

Verify: run the connection check. Embrasure should list the selected scope without receiving permission to write warehouse data.

Sync the catalog

Start a catalog sync, then open Catalog and find a known table. Confirm its connector, fully qualified name, columns, and schema before evaluating lineage.

Verify: the table identity matches the physical object you expect.

Add lineage evidence

Choose one or more paths:

• Warehouse-native history for BigQuery, Snowflake, Databricks, Redshift, and supported SQL sources.
• OpenLineage for Airflow, Spark, Flink, Dataproc, EMR, Glue jobs, and custom applications.
• Artifact imports for dbt, SQLMesh, DataHub, and Fivetran.
• Cloud lineage feeds from Google Data Lineage, Amazon DataZone, or Microsoft Purview.

Verify: run a job that reads one cataloged asset and writes another, then wait for asynchronous processing.

Check impact

npm install --global @embrasure/cli
embrasure auth login
embrasure doctor
embrasure lineage impact catalog.schema.table \
  --connector YOUR_CONNECTOR_ID \
  --direction both

An empty graph can mean no evidence arrived, an endpoint did not resolve, or request bounds hid part of the graph. Follow coverage and troubleshooting before treating absence as proof.

Put lineage to work

Arm a Scout, install change review, or connect an automation surface. Each workflow uses the same resolved catalog and evidence-backed graph.