Lineage

Lineage provides a unified view of how data moves across systems, from raw sources to transformed datasets, models, and downstream reports. It helps users understand dependencies, trace data transformations, and analyze the impact of changes at any point in the data flow.

By visualizing both upstream and downstream relationships, Lineage enables teams to troubleshoot faster, maintain trust in data, and ensure reliability across their analytics ecosystem.

Why Lineage Matters

• End-to-End Visibility: See where data originates, how it is transformed, and where it’s consumed.
• Impact Analysis: Identify which assets (reports, dashboards, or datasets) are affected by upstream failures or schema changes.
• Column-Level Insight: Trace data lineage down to individual columns, with visibility into data types and quality scores.
• Faster Troubleshooting: Quickly pinpoint root causes and prioritize remediation based on downstream impact.

Types of Lineage Creation

Lineage in the platform can be created in two ways: Manual Lineage and Dynamic Lineage.

Manual Lineage

Manual lineage allows users to define and build lineage paths manually through the UI or APIs. This approach is useful when:

• The source systems do not automatically support dynamic lineage.
• You need to define custom relationships between assets that are not connected through built-in integrations.
• Teams want to enrich existing lineage with additional business logic or external data flow links.

Manual lineage can be created by:

• Using the Add Lineage option in the Lineage view to link assets.
• Leveraging public APIs to define relationships programmatically between upstream and downstream assets.

This gives teams full flexibility to model complex or proprietary data flows that go beyond automated system connections.

Dynamic Lineage

Dynamic lineage is automatically derived from supported source connections through asset fingerprinting and metadata analysis. When the platform connects to compatible systems (for example, Snowflake, BigQuery, Databricks, Power BI, Tableau, and others), it analyzes the metadata to infer relationships across systems.

Imprint-Based Asset Inference for Power BI Assets

Dynamic lineage for Power BI relies on imprint-based asset inference, which connects Power BI assets with underlying physical data sources such as Snowflake or BigQuery.

Here’s how it works:

• Crawling of Snowflake or other physical data sources generates imprints in the format native to each source type.
• Crawling of Power BI assets generates additional custom metadata for:
• Power BI Semantic Model Tables and Columns
• Power BI Dataflow Entities and Attributes
• When the Power BI crawler finishes, an imprint matching job is triggered asynchronously. This job matches Power BI imprints against Snowflake or other physical data source assets to infer lineage automatically.

This imprint-based approach ensures that lineage between Power BI and its data sources is created dynamically, accurately mapping transformations and relationships down to the column level.

Lineage Views

Lineage includes two complementary views that serve different purposes: Lineage and Relationships.

Lineage Tab

The Lineage tab focuses on the movement of data across connected systems and assets.

Key capabilities include:

• Directed flow graph showing upstream and downstream paths.
• Column-level tracing across datasets, models, and reports.
• Contextual details like column names, data types, and quality metrics.
• Impact Analysis visualization — highlighting affected downstream assets when upstream issues occur.

This view answers questions like:

“Where does this data come from?” “Which reports depend on this dataset?” “What will break if this table fails?”

Cross-System Lineage

The platform supports end-to-end, column-level lineage across multiple systems. This includes visibility into how data flows between BI tools, transformation layers, and data warehouses.

Supported examples:

• Power BI → Snowflake / Databricks / BigQuery / other supported sources
• Tableau → Snowflake / Databricks / BigQuery / other supported sources

Lineage across these platforms is derived using DBT model mappings and imprint-based inference, providing a complete picture of how data moves from source to visualization.

Key benefits:

• Trace BI fields all the way back to their source columns.
• View transformation logic, relationships, and calculated field expressions (such as DAX or Tableau Calculations).
• Identify the downstream impact of data changes across systems.
• Build confidence in report accuracy with detailed, system-level metadata mappings.

Relationships Tab

The Relationships Tab provides a high-level, structural view of how assets are organized and logically related. Unlike the Lineage Tab, which focuses on data flow, the Relationships Tab helps users understand how assets are grouped, owned, and connected within the platform.

Key capabilities include:

• Displays parent-child hierarchies, such as:
• Workspace → Report → Semantic Model → Dataset
• Project → Schema → Table → Column
• Shows ownership and logical grouping of assets within domains, projects, or workspaces.
• Helps users navigate the data structure in large environments where lineage graphs may be extensive.
• Provides insight into organizational and contextual relationships, complementing the data flow view.

Use the Relationships Tab to answer questions like:

“Which workspace or project does this asset belong to?” “Which reports and dashboards are grouped under a workspace?” “How are datasets and assets organized across domains?”

Accessing Lineage

1. Navigate to Discover Assets under the Reliability section in the navigation menu.
2. Use the search bar to find a data asset by name or data source.
3. Select the asset to open its details.
4. Choose the Lineage tab to explore data flow or the Relationships tab for structural context.

Working with the Lineage View

In the Lineage view, you can:

• Visualize data flow between systems and downstream assets.
• Explore column-level lineage with names, data types, and quality scores.
• Search within tables or columns using the built-in search bar.
• Add custom or missing lineage connections.
• View asset metadata, related assets, and dependency trees via the Relationships tab.