Data Observability

Data Lineage

Visualize the complete flow of your data. From sources to destinations, at table and column level. Impact analysis before making changes. Graph snapshots and health monitoring.

Table Lineage Graph

Exploring: gold.customer_360 - Upstream and downstream dependencies

Upstream
Selected
Downstream
Tables tracked:247
Column mappings:1,842
PII fields:156
PHI fields:89

Lineage Capabilities

Table-Level Lineage

Track upstream sources and downstream consumers with configurable depth (1-10 levels). Multi-node type support for tables, columns, and pipelines.

Column-Level Lineage

Track column-level transformations with confidence scoring. Direct copy detection, transformation expressions, and batch lineage creation.

Multi-Dialect SQL Parsing

Automatic lineage from SQL with 13+ dialects: Spark, Databricks, PostgreSQL, BigQuery, Snowflake, Redshift, dbt, and more via sqlglot.

Impact Analysis

Analyze downstream impact with severity levels (low → critical). Configurable depth (3-20 levels), AI recommendations, and full analysis history.

Graph Snapshots

Save point-in-time lineage views. Load, compare, and manage snapshots with metadata tracking. One-click save from canvas.

Health Monitoring

Lineage coverage metrics with status indicators. Stale lineage detection, missing lineage alerts, and hub table identification.

Interactive Canvas

React Flow visualization with mini-map, animated edges, zoom controls. Custom nodes for tables, columns, and pipelines with status indicators.

OpenLineage Standard

Consume events from Airflow, Spark, dbt, Dagster, Flink. Emit to Marquez, DataHub, Atlan, Collibra, Purview, OpenMetadata.

Pipeline Integration

Auto-detect lineage from pipeline canvas. Link lineage to source pipelines and runs with status visibility (active/paused/error).

Column-Level Transformation Tracking

Track exactly how each column is derived. Color-coded by transformation type.

Tracked Transformations

direct_copy

Direct column copy

cast

Type conversion

concat

String concatenation

aggregate

SUM, COUNT, AVG, etc.

join

Multi-table join

filter

WHERE conditions

hash

SHA-256 hashing

transform

Custom expression

derived

Derived columns

unknown

Custom logic

Example: email_hash derivation

Source Column

customers.email

VARCHAR(255)

Transformation: hash

SHA256(LOWER(TRIM(email)))

PII Protection applied

Target Column

customer_analytics.email_hash

VARCHAR(64)

Before You Change

Impact Analysis

Before making schema changes, understand exactly what will be affected. Nexion analyzes downstream dependencies and provides severity ratings.

  • Identify all impacted tables
  • Identify affected columns
  • List pipelines that will break
  • Severity rating (low to critical)
  • Recommended actions

Impact Analysis Result

HIGH

Change Proposed

Drop column: customers.phone

Impacted Tables (3)

• customer_360

• marketing_contacts

• support_tickets

Affected Pipelines (2)

• daily_customer_sync

• marketing_etl

Recommendations

• Update customer_360 pipeline first

• Notify marketing team

• Schedule change for off-hours

Open Standard

OpenLineage Compatible

Nexion emits lineage events in OpenLineage format, the open standard from the Linux Foundation. Integrate with your existing data governance stack.

OpenLineage Standard

OpenLineage is the open standard for data lineage capture and emission. Adopted by Airflow, Spark, dbt, Flink, and Dagster.

Evento START/RUNNING/COMPLETE/FAIL
Input/Output datasets con schema
Column-level lineage facets
SQL job facets

Compatible Tools

Send Nexion events to your preferred data governance tool:

Marquez

Open source (LF AI)

DataHub

LinkedIn / Acryl Data

Atlan

Data Catalog

Collibra

Enterprise Governance

Purview

Microsoft Azure

OpenMetadata

Open source

Lineage Architecture

Nexion Pipelines

Ingest, Transform, Load

Nexion Lineage

PostgreSQL + sqlglot

OpenLineage Events

Standard format

Your Catalog

Marquez, DataHub, etc.

Native lineage in Nexion + optional emission to external tools via OpenLineage

Visualize your data flow

See the complete lineage of your data estate with Nexion.