# Foundational Data Lineage

> Data control plane that builds a knowledge graph of code and data movement, with column-level lineage from source code, pull request impact analysis, and data contracts.

Source: https://toolsfordata.com/tools/foundational/ · Updated 2026-11-19

[Website](https://www.foundational.io/) · [Docs](https://docs.foundational.io/) · [Pricing](https://www.foundational.io/pricing)

Best for: data and engineering teams that want to see the downstream impact of code changes in Git pull requests and govern data across repositories, databases, and BI.

## Overview

- Core Strength: Column-level lineage from source code (Python, Java, Scala, .NET, COBOL, SQL), with pull request impact checks and data contracts in Git
- Lineage Depth: Full — automated, table, column, end-to-end…
- Catalog Curation: Partial — certification, usage stats
- Stewardship Workflows: Partial — review/approval
- Policy & Classification: Not supported
- Compliance Depth: Partial — audit trail
- Marketplace & Data Products: Partial — subscriptions
- Built-in Data Quality: Partial — SQL tests, profiling, anomaly detection, contracts
- AI & Agent Interface: Partial — MCP, chat, AI docs, knowledge graph
- AI Governance: Partial — AI lineage
- Developer Surface: Partial — API, dbt
- Deployment Model: Partial — SaaS, hybrid
- Security & Access: Full — SSO, SCIM, RBAC, audit log…
- Notable Caveat: A lineage and change-control tool, not a full catalog: no glossary, access policies or marketplace

## What it's used for

Foundational is used to see the downstream impact of code changes before they merge: it builds column-level lineage from source code and posts the impact to pull requests, with data contracts as YAML files next to the code.

## Integrations

Foundational reads code from GitHub, GitLab, Bitbucket and Azure DevOps to trace lineage across Snowflake, Databricks, BigQuery, Redshift, dbt and Airflow, and works with 11 AI agent clients.

## Review

Strengths:
- Lineage comes from source code, so impact shows before the code runs.
- Posts downstream impact to pull requests.
- Data contracts are YAML files next to the code, turned into monitors.

Caveats:
- Needs access to your Git repositories to parse code.
- Lineage covers the languages it parses: SQL, Spark, Python and Java.

Recommendation: For field-level lineage of sensitive data across services, consider Gable. For data diffs on dbt pull requests only, Recce; for a full catalog with lineage, Atlan; for open-source runtime lineage, OpenLineage.

## Features

### Data Catalog

- Automated Metadata Harvesting: Yes — Connectors scan warehouses, BI tools and ETL on a schedule, and code repositories on every change. ([source](https://docs.foundational.io/en/articles/12802881-solution-architecture))
- Search & Discovery: Yes — Search with filters by keyword, asset type, platform, database, schema and certification status. ([source](https://docs.foundational.io/en/articles/14044315-use-the-data-catalog))
- AI-Powered Search: Unknown — Foundational IQ answers questions in plain English, but semantic search in the catalog is not documented. ([source](https://docs.foundational.io/en/articles/14119487-what-can-foundational-iq-do-for-you))
- Usage & Popularity Statistics: Yes — Tracks query frequency, access patterns and user activity per asset, with historical trends, to find unused tables and dashboards. ([source](https://docs.foundational.io/en/articles/14044890-track-asset-usage-and-costs))
- Custom Fields & Templates: Unknown — Manual lineage uploads accept custom attributes, but custom metadata fields in the catalog are not documented. ([source](https://docs.foundational.io/en/articles/15443234-how-to-upload-manual-lineage-xlsx))
- Asset Certification: Yes — Owners can set a certification status on assets, such as certified or deprecated, and filter by it. ([source](https://docs.foundational.io/en/articles/14044714-maintain-data-asset-documentation))
- Business Glossary: No — No business glossary is documented. ([source](https://docs.foundational.io/en/articles/14044315-use-the-data-catalog))
- Metadata Propagation: Unknown — The Lineage API docs suggest propagating PII labels downstream as an automation you build; no built-in propagation is documented. ([source](https://docs.foundational.io/en/articles/10067204-get-started-with-the-lineage-api))

### Data Lineage

- Table-Level Lineage: Yes — Lineage graph of tables, files, jobs and dashboards across the stack. ([source](https://docs.foundational.io/en/articles/12802881-solution-architecture))
- Column-Level Lineage: Yes — Column-level lineage from SQL, dbt, Spark and application code (Java, Scala, .NET, Python, COBOL), without running the code. ([source](https://docs.foundational.io/en/articles/9479381-sql-lineage))
- End-to-End Lineage: Yes — From upstream operational databases and application code through warehouses to BI dashboards such as Tableau, Looker and Power BI. ([source](https://docs.foundational.io/en/articles/12802881-solution-architecture))
- Impact Analysis: Yes — Every pull request is compared against the lineage graph to show downstream impact before merge, and can block the merge. ([source](https://docs.foundational.io/en/articles/14047433-assess-impact-before-merging-code-changes))
- Automated Lineage Extraction: Yes — Static analysis of SQL and application code (Python, Java, Scala, .NET, COBOL, ORMs), plus connectors for dbt, Airflow, warehouses and BI tools. ([source](https://docs.foundational.io/en/articles/15082464-supported-languages-and-platforms))
- OpenLineage Support: Yes — Ingests OpenLineage events and exports lineage as OpenLineage events to Amazon S3, which Collibra can read. ([source](https://docs.foundational.io/en/articles/18328197-export-lineage-as-openlineage-to-amazon-s3))

### Data Stewardship

- Critical Data Elements: No — No register of critical data elements is documented. ([source](https://docs.foundational.io/en/articles/14044315-use-the-data-catalog))
- Review & Approval Workflows: Yes — Workflows and Alerts assign reviewers automatically to changes that modify tables, inside the pull request flow. ([source](https://docs.foundational.io/en/articles/10067204-get-started-with-the-lineage-api))
- Governance Campaigns: No — No bulk governance campaigns are documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- Governance Scorecards: No — No governance scorecards are documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- Metadata Automation: Yes — Workflows and Alerts automate notifications and reviewer assignment when specific changes or events occur. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))

### Data Classification

- Sensitive Data Discovery: No — Foundational reads metadata and code only, never data; sensitive data findings come from the Cyera integration. ([source](https://docs.foundational.io/en/articles/17884079-how-to-set-up-cyera))
- Data Policy: No — Pull request checks can block moving classified columns to unapproved targets, but access to data is not governed. ([source](https://docs.foundational.io/en/articles/17884079-how-to-set-up-cyera))
- Native Policy Enforcement: No — No access policies are pushed to data platforms. ([source](https://docs.foundational.io/en/articles/17884079-how-to-set-up-cyera))
- Data Masking: No — No masking; Foundational never reads data. ([source](https://docs.foundational.io/en/articles/9515541-security-and-compliance))

### AI Features

- AI Chat Assistant: Yes — Foundational IQ answers questions about lineage, costs and sensitive data in plain English and can open pull requests with your approval. ([source](https://docs.foundational.io/en/articles/14119487-what-can-foundational-iq-do-for-you))
- AI Documentation Generation: Yes — AI fills documentation gaps from lineage, usage, source code and neighboring assets, and updates it when the stack changes. ([source](https://docs.foundational.io/en/articles/14044714-maintain-data-asset-documentation))
- Agentic Metadata Enrichment: Unknown — Documentation is generated and updated automatically, but autonomous tagging or classification by agents is not documented. ([source](https://docs.foundational.io/en/articles/14044714-maintain-data-asset-documentation))

### Data Compliance

- Audit Trail: Yes — The Lineage API returns the graph as it existed at a past point in time, and Foundational IQ uses the change history of code and assets. ([source](https://docs.foundational.io/en/articles/10067204-get-started-with-the-lineage-api))
- Compliance Reporting: No — No compliance reports are documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- Regulatory Framework Mapping: No — No regulation templates; BCBS 239 is only discussed as a reason for mainframe lineage. ([source](https://docs.foundational.io/en/articles/16189006-mainframe-cobol-lineage-on-aws-carddemo))
- Data Activity Monitoring: No — Query history is read for usage and costs, not monitored for access to sensitive data. ([source](https://docs.foundational.io/en/articles/14044890-track-asset-usage-and-costs))
- ROPA: No — No GDPR Article 30 register is documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- Data Retention Policies: No — No retention rules are documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- Data Subject Rights: No — No data subject request workflow; Foundational does not access data. ([source](https://docs.foundational.io/en/articles/9515541-security-and-compliance))

### Collaboration

- Alert Notifications: Yes — Alerts to Slack, Microsoft Teams and email when changes affect downstream assets or contracts are violated. ([source](https://docs.foundational.io/en/articles/13959754-data-contracts-how-foundational-simplifies-automation-and-enforcement))
- Discussions & Annotations: No — No comments or discussions on catalog assets are documented. ([source](https://docs.foundational.io/en/articles/14044315-use-the-data-catalog))
- Ticketing Integrations: No — No Jira or ServiceNow integration is documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))

### Data Marketplace

- Self-Service Access Requests: No — No access requests. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- Data Products: No — No data products are documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- Data Mesh Support: No — Contracts live in each team's repository, but domains and data products are not modelled. ([source](https://docs.foundational.io/en/articles/13959754-data-contracts-how-foundational-simplifies-automation-and-enforcement))
- Subscription Management: Yes — Workflows notify chosen people when changes affect specific tables or dashboards. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))

### Data Quality

- SQL Table Tests: Yes — Contract business rules check required fields, value ranges and key uniqueness. ([source](https://docs.foundational.io/en/articles/14587381-monitor-data-contracts))
- Data Health Scores: Unknown — Not documented. ([source](https://docs.foundational.io/en/articles/14044315-use-the-data-catalog))
- Anomaly Detection: Yes — Observability monitors catch freshness and volume problems, with anomaly rules and custom thresholds. ([source](https://docs.foundational.io/en/articles/15700888-what-foundational-detects-issue-detection-catalog))
- Data Profiling: Yes — Table statistics such as null counts, zero counts and distinct values per column, with custom thresholds; lighter than dedicated profilers. ([source](https://docs.foundational.io/en/articles/10943880-configuring-custom-thresholds-for-observability-tests))
- Data Contracts: Yes — YAML contracts (.fd.yml) next to the code, with schema, business, anomaly and SLA rules, checked on pull requests and continuously. ([source](https://docs.foundational.io/en/articles/14587381-monitor-data-contracts))

### AI Governance

- AI Asset Inventory: No — No catalog of AI models or agents is documented. ([source](https://docs.foundational.io/en/articles/14044315-use-the-data-catalog))
- AI Lineage: Yes — Code analysis traces data from warehouses to external AI endpoints such as OpenAI and to downstream ML models (training-data lineage). ([source](https://docs.foundational.io/en/articles/14640995-extracting-lineage-from-python))
- AI Regulatory Compliance: No — Not documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- AI Model Health Score: No — Not documented. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))
- AI Agent Access Governance: No — The MCP server gives agents lineage context, but nothing is enforced at query time. ([source](https://docs.foundational.io/en/articles/11194236-use-foundational-mcp-with-your-ai-tools))

### Master Data Management

- Master Data Management: No — No MDM. ([source](https://docs.foundational.io/en/articles/12800446-navigate-the-interface))

### Workflows

- REST / GraphQL API: Yes — REST Lineage API to query entities and the lineage graph, including as it existed at a past point in time. ([source](https://docs.foundational.io/en/articles/10067204-get-started-with-the-lineage-api))
- dbt Integration: Yes — Parses dbt and dbt Fusion projects for column-level lineage and checks dbt changes on pull requests. ([source](https://docs.foundational.io/en/articles/15082464-supported-languages-and-platforms))
- Python SDK: No — No SDK; a sample Python project calls the REST API. ([source](https://docs.foundational.io/en/articles/10067204-get-started-with-the-lineage-api))
- Webhooks: Unknown — Teams notifications use an incoming webhook; general outbound webhooks are not documented. ([source](https://docs.foundational.io/en/articles/11533006-how-to-set-up-microsoft-teams-notifications))

### Data Context Layer

- MCP Server: Yes — MCP server for Claude, ChatGPT, Copilot, Cursor, Snowflake Cortex and others, with lineage, usage and code definitions. ([source](https://docs.foundational.io/en/articles/11194236-use-foundational-mcp-with-your-ai-tools))
- Knowledge Graph: Yes — One directed graph of code, tables, columns, jobs and dashboards with owners and tags, queried by the API, MCP and Foundational IQ. ([source](https://docs.foundational.io/en/articles/10067204-get-started-with-the-lineage-api))
- Semantic Models: No — No semantic model generation is documented. ([source](https://docs.foundational.io/en/articles/14119487-what-can-foundational-iq-do-for-you))
- Verified Queries: No — Not documented. ([source](https://docs.foundational.io/en/articles/14119487-what-can-foundational-iq-do-for-you))
- Semantic Layer Interop: Unknown — LookML and dbt are parsed for lineage; importing metric definitions is not documented. ([source](https://docs.foundational.io/en/articles/15082464-supported-languages-and-platforms))
- Unstructured Data Catalog: No — Lineage can reach unstructured data lakes, but documents are not catalogued. ([source](https://docs.foundational.io/en/articles/14640995-extracting-lineage-from-python))

### Security

- SSO: Yes — SSO with SAML 2.0 or OpenID Connect. ([source](https://docs.foundational.io/en/articles/12807462-set-up-sso-saml-and-oidc))
- RBAC: Yes — Two user roles, Admin and Read Only, plus API token roles. ([source](https://docs.foundational.io/en/articles/12807799-user-management))
- Audit Log: Yes — Audit Logs record user actions, login events and account changes. ([source](https://docs.foundational.io/en/articles/12807799-user-management))
- Compliance Certifications: Yes — SOC 2 Type II. ([source](https://docs.foundational.io/en/articles/9515541-security-and-compliance))
- SCIM Provisioning: Yes — SCIM provisioning from Okta, Azure and other identity providers. ([source](https://docs.foundational.io/en/articles/12807578-set-up-scim-provisioning))

### Deployment

- SaaS Deployment: Yes — Cloud platform that reads metadata and code, never data. ([source](https://docs.foundational.io/en/articles/9515541-security-and-compliance))
- Self-Hosted Deployment: No — Even with local processing, the lineage graph is synced to Foundational's cloud. ([source](https://docs.foundational.io/en/articles/13105912-foundational-on-premise-deployments))
- Hybrid Deployment: Yes — Enterprise plans run an agent or the parsing and analysis inside your network; only the sanitized lineage graph is synced. ([source](https://docs.foundational.io/en/articles/13105912-foundational-on-premise-deployments))
- Data Residency: Unknown — Multi-region deployments are mentioned, but a region choice for the SaaS is not documented. ([source](https://docs.foundational.io/en/articles/12802881-solution-architecture))

### Open Source

- Open Source: No — Proprietary. ([source](https://www.foundational.io/pricing))

## Integrations

95 integrations, each sourced from the vendor docs.

- AI Agents: ChatGPT, Claude, Claude Code, Cursor, Devin, Factory Droid, GitHub Copilot, Microsoft Copilot, OpenAI Codex, Salesforce Agentforce, Snowflake Cortex
- Authentication: Google OAuth, Google Workspace, JumpCloud, Microsoft Entra, Okta, Onelogin, OpenID, Ping, Rippling, SAML
- BI: Amazon QuickSight, Apache Superset, Domo, Hex, Looker, Looker Studio, Metabase, Mode, Power BI, SAP Business Objects, Sigma, Sisense, Spotfire, Tableau, ThoughtSpot
- Code Repositories: Azure DevOps, Bitbucket, GitHub, GitLab
- Data Applications: HubSpot, SAP ERP, Salesforce, Workday
- Data Governance: Collibra, OpenLineage
- Data Integration: AWS Glue, Airbyte, Azure Data Factory, Fivetran, Google Datastream
- Data Orchestration: Apache Airflow
- Data Platform: Alteryx, IBM DataStage, Informatica, Microsoft Fabric, TIBCO Data Virtualization, Talend
- Data Security: Cyera
- Data Transformation: Apache Spark, Coalesce, Google Dataflow, Google Dataform, Matillion, SQLMesh, dbt
- Data Warehouse: Amazon Redshift, Databricks SQL, Google BigQuery, IBM DB2, OpenText Analytics Database (Vertica), Oracle DB, Snowflake
- Databases: Amazon Aurora, Amazon RDS, Apache Cassandra, Apache CouchDB, Azure SQL Database, Couchbase, Google BigTable, Google Cloud Spanner, Greenplum, Mainframe, MariaDB, Microsoft SQL Server, MongoDB, MySQL, PostgreSQL
- Notifications: Microsoft Teams, Slack
- Query Engine: Amazon Athena, Hive SQL, Trino
- Streaming: Apache Kafka, Confluent

## Alternatives

- [Gable](https://toolsfordata.com/tools/gable/): The closest match: lineage from static code analysis, aimed at sensitive data flows and compliance.
- [Recce](https://toolsfordata.com/tools/recce/): Consider Recce if your pull requests are dbt changes: data diffs and impact before merge.
- [Solidatus](https://toolsfordata.com/tools/solidatus/): Consider Solidatus for lineage as visual dependency graphs, built for audits in financial services.
- [Atlan](https://toolsfordata.com/tools/atlan/): Consider Atlan for a catalog with column-level lineage and data contracts.
- [OpenLineage](https://toolsfordata.com/tools/openlineage/): The free route: collect lineage events from Airflow, Spark, dbt, Flink and Dagster with the open standard, then pick a backend.
