# Gable Data Lineage

> Field-level lineage from static analysis of Java and TypeScript service code, with PII tagging, versioned lineage approvals, and exports for audit evidence.

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

[Website](https://www.gable.ai/) · [Docs](https://docs.gable.ai/intro)

Best for: regulated enterprises that need to answer compliance, privacy, and model risk questions about where sensitive data flows, without runtime agents.

## Overview

- Core Strength: Field-level lineage from application code (Java, TypeScript), not just the warehouse
- Lineage Depth: Partial — automated, table, column, impact analysis
- Catalog Curation: Partial — propagation
- Stewardship Workflows: Partial — review/approval
- Policy & Classification: Partial — sensitive-data discovery
- Compliance Depth: Partial — audit trail, reporting
- Marketplace & Data Products: Not supported
- Built-in Data Quality: Partial — contracts
- AI & Agent Interface: Partial — AI docs
- AI Governance: Not confirmed
- Developer Surface: Partial — API, webhooks
- Deployment Model: Partial — SaaS, hybrid
- Security & Access: Partial — SSO, RBAC, certifications
- Notable Caveat: Lineage and audit evidence only, and its current docs no longer cover data contracts

## What it's used for

Gable reads application source code to map field-level data flows across services and data stores. Teams in regulated industries use it to answer where sensitive data comes from and goes, and to see what a code change affects.

## Lineage

Gable builds field-level lineage from Java and TypeScript service code. Each mapping shows whether a value is passed through, calculated or redacted, with each step linked to the line in GitHub.

## Compliance

Evidence comes from lineage: trace a regulated field, export it to CSV or the API, and tie it to the commit. Approvals sign off a lineage version as the baseline.

## Classification

Scans tag PII fields automatically; other tags are applied by hand and follow the field everywhere it appears.

## Quality

Gable's website still describes contracts drafted from code, but its current docs (June 2026) don't cover them, and Gable runs no tests on data.

## Pricing

Gable doesn't publish prices and has no pricing page. Sales is demo-only.

## Deployment

Gable runs as single-tenant SaaS: each customer gets its own AWS account. The CLI scans code in your CI/CD and sends only code metadata.

## Security

Gable supports SAML SSO and three roles, with a SOC 2 Type 2 audit from June 2025.

## Integrations

Gable reads application code in GitHub and runs in GitHub Actions, with no warehouse or database connectors: it catches schema changes before they ship.

## Review

Strengths:
- Lineage comes from source code, so it covers application services that warehouse catalogs never see.
- Field-level mappings show the kind of transformation and link each step to the exact line of code.
- Lineage is versioned per commit with approvals, a point-in-time record for auditors.

Caveats:
- Static analysis supports only Java and TypeScript; gaps need manual edges and overrides.
- No catalog, glossary, stewardship, access policies or data quality testing.
- Gable moved from data contracts to compliance lineage: its current docs no longer cover contracts, and merge impact reports are still coming soon.

Recommendation: Choose Gable if regulators ask how data moves through your application code, not just your warehouse. For lineage from the warehouse to dashboards, use a catalog.

## Facts

- Gable was founded by Chad Sanderson, who led data at Convoy, and raised a $7M seed in September 2023.
- Gable raised a $20M Series A in April 2025, with Databricks Ventures and Capital One Ventures.
- Gable moved its docs to docs.gable.ai in June 2026; they now cover code lineage and compliance evidence.

## Pricing

- Model: custom
- Entry team price: Custom
- Entry enterprise price: Custom
- Source: https://www.gable.ai/ (checked 2026-11-12)

## Features

### Data Catalog

- Automated Metadata Harvesting: Yes — Scanning a service's source code creates its code component automatically, with ingresses, egresses and field mappings; data stores are added by hand or published from Avro and Protobuf files. ([source](https://docs.gable.ai/docs/onboarding/services-and-data-components))
- Search & Discovery: No — Search is limited to finding fields and components in the Lineage Explorer; there is no faceted asset search by owner, tag or certification. ([source](https://docs.gable.ai/docs/tracing/end-to-end-lineage))
- AI-Powered Search: No — No natural-language or semantic search is documented; field search in the Lineage Explorer is by name. ([source](https://docs.gable.ai/docs/tracing/end-to-end-lineage))
- Usage & Popularity Statistics: No — Gable reads code, not query logs, so it shows no query counts or popularity per asset. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- Custom Fields & Templates: No — Data store components take optional freeform metadata, but there are no typed custom fields or templates per asset type. ([source](https://docs.gable.ai/docs/onboarding/services-and-data-components))
- Asset Certification: No — Approvals sign off a component's lineage version as a baseline, which is not a trust label on tables or dashboards. ([source](https://docs.gable.ai/docs/impact/approvals))
- Business Glossary: No — There is no glossary module; friendly names on fields are the only business-facing labels. ([source](https://docs.gable.ai/docs/tracing/end-to-end-lineage))
- Metadata Propagation: Yes — A tag attaches to the field's underlying address, so tagging a field once carries the tag across every mapping that field feeds. ([source](https://docs.gable.ai/docs/tracing/data-tagging))

### Data Lineage

- Table-Level Lineage: Yes — The Lineage Explorer graph shows services and data stores (tables, Kafka streams, files) connected by cross-service edges. ([source](https://docs.gable.ai/docs/tracing/end-to-end-lineage))
- Column-Level Lineage: Yes — Each field mapping records the ingress and egress field, whether the value is passed through, calculated or redacted, and the code path steps with file and line. ([source](https://docs.gable.ai/docs/tracing/field-level-lineage-metadata))
- End-to-End Lineage: No — Lineage spans application services and data stores, but no BI tool or dashboard integration is documented, so it does not reach reports. ([source](https://docs.gable.ai/docs/tracing/end-to-end-lineage))
- Impact Analysis: Yes — Version comparison classifies changes as breaking, behavior change or safe addition, and each field shows its downstream consumers; per-merge impact reports in CI/CD are still marked coming soon. ([source](https://docs.gable.ai/docs/impact/version-changes))
- Automated Lineage Extraction: Yes — Static analysis of application source code (run in CI or with the Gable CLI) extracts field-level lineage from ingress to egress, chained across services and data stores. ([source](https://docs.gable.ai/docs/tracing/end-to-end-lineage))
- OpenLineage Support: No — Lineage comes from Gable's own static code analysis and lineage events API; no OpenLineage emit or ingest is documented. ([source](https://docs.gable.ai/api-reference/introduction))

### Data Stewardship

- Critical Data Elements: No — No way to register critical data elements is documented; regulated fields are traced by search or tags. ([source](https://docs.gable.ai/docs/tracing/compliance-evidence))
- Review & Approval Workflows: Yes — Reviewers step through a lineage version's changes and approve it with a note; later overrides mark it as changed since approval for re-approval. ([source](https://docs.gable.ai/docs/impact/approvals))
- Governance Campaigns: No — No bulk governance campaigns are documented. ([source](https://docs.gable.ai/intro))
- Governance Scorecards: No — No maturity scorecards are documented; the Lineage quality tab measures lineage coverage, not governance maturity. ([source](https://docs.gable.ai/docs/onboarding/coverage-and-completeness))
- Metadata Automation: Yes — Actions trigger on data asset events (created, field added, changed or removed) by source or namespace and send to Slack, email or a webhook. ([source](https://docs.gable.ai/api-reference/action/create-a-new-action))

### Data Classification

- Sensitive Data Discovery: Yes — Each scan applies a system PII tag to fields that look like emails, social security numbers or card numbers; PHI, PIFI and Secret tags are applied by hand. ([source](https://docs.gable.ai/docs/tracing/data-tagging))
- Data Policy: No — Gable does not control who can read or query data; it maps where data flows. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- Native Policy Enforcement: No — Gable pushes no access or masking policies to data platforms. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- Data Masking: No — Gable detects where code redacts or masks a value (the Redaction computation type) but does not mask data itself. ([source](https://docs.gable.ai/docs/tracing/field-level-lineage-metadata))

### AI Features

- AI Chat Assistant: No — No chat assistant is documented. ([source](https://docs.gable.ai/intro))
- AI Documentation Generation: Yes — Code path steps include a human-readable summary of the transformation at each step, and changelog event titles are LLM-generated. ([source](https://docs.gable.ai/docs/tracing/field-level-lineage-metadata))
- Agentic Metadata Enrichment: No — No agents that tag or document assets on their own are documented; only PII tagging is automatic. ([source](https://docs.gable.ai/docs/tracing/data-tagging))

### Data Compliance

- Audit Trail: Yes — The Scan Log records every analyzed lineage version with branch, commit and timestamps, and approvals record who approved, when and why. ([source](https://docs.gable.ai/docs/tracing/compliance-evidence))
- Compliance Reporting: Yes — Field lineage exports to CSV hop by hop, and the Lineage Export API returns a run's full components, schemas and edges for audit evidence. ([source](https://docs.gable.ai/docs/tracing/exporting-lineage))
- Regulatory Framework Mapping: No — Gable publishes BCBS 239 content, but the docs describe no mapping of assets or controls to regulations. ([source](https://docs.gable.ai/docs/tracing/compliance-evidence))
- Data Activity Monitoring: No — Gable analyzes code statically in CI/CD and does not monitor access to data sources. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- ROPA: No — No GDPR Article 30 processing register is documented. ([source](https://docs.gable.ai/docs/tracing/compliance-evidence))
- Data Retention Policies: No — No retention or deletion rules are documented. ([source](https://docs.gable.ai/docs/tracing/compliance-evidence))
- Data Subject Rights: No — No DSAR or deletion request handling is documented. ([source](https://docs.gable.ai/docs/tracing/compliance-evidence))

### Collaboration

- Alert Notifications: Yes — Actions send data asset change events to Slack channels, email addresses or webhooks. ([source](https://docs.gable.ai/api-reference/action/create-a-new-action))
- Discussions & Annotations: No — Field mappings take notes, but there are no threaded questions or comments on assets. ([source](https://docs.gable.ai/docs/onboarding/field-mappings-and-overrides))
- Ticketing Integrations: No — No Jira or ServiceNow integration is documented; defect reports go to Gable. ([source](https://docs.gable.ai/docs/onboarding/field-mappings-and-overrides))

### Data Marketplace

- Self-Service Access Requests: No — No access request flow is documented. ([source](https://docs.gable.ai/docs/availability_security_compliance/role_based_access_control))
- Data Products: No — No data product publishing is documented. ([source](https://docs.gable.ai/intro))
- Data Mesh Support: No — No domain workspaces or federated data product model is documented; namespaces only scope scans. ([source](https://docs.gable.ai/docs/cli/introduction))
- Subscription Management: No — No subscriptions to data products are documented; actions notify chosen Slack channels, emails or webhooks. ([source](https://docs.gable.ai/api-reference/action/create-a-new-action))

### Data Quality

- SQL Table Tests: No — Gable reads code, not data, so it runs no SQL tests on tables. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- Data Health Scores: No — No data health score is documented. ([source](https://docs.gable.ai/intro))
- Anomaly Detection: No — Gable does not read data values, so there is no anomaly detection. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- Data Profiling: No — Gable does not read data values, so there is no profiling. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- Data Contracts: Yes — Gable drafts data contracts from existing data code and flags code changes that would violate them; the current docs cover lineage only, though the API still records contract violation events. ([source](https://www.gable.ai/overview/product-overview))

### AI Governance

- AI Asset Inventory: No — The agent project scan discovers LangChain agents, tools and models in a repository, but Gable keeps no registry of models with owners, versions and purpose. ([source](https://docs.gable.ai/docs/settings/docker_image))
- AI Lineage: Unknown — Only the website mentions model input lineage (which fields reach ML models and chatbots); no docs page covers it. ([source](https://docs.gable.ai/docs/tracing/compliance-evidence))
- AI Regulatory Compliance: No — No EU AI Act, NIST AI RMF or ISO 42001 controls are documented. ([source](https://docs.gable.ai/intro))
- AI Model Health Score: No — No model trust score is documented. ([source](https://docs.gable.ai/intro))
- AI Agent Access Governance: No — Gable does not sit between agents and data at query time. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))

### Master Data Management

- Master Data Management: No — Gable has no MDM. ([source](https://docs.gable.ai/intro))

### Workflows

- REST / GraphQL API: Yes — The REST API covers components, lineage export, cross-service edges, tags, approvals, actions and webhooks, with a downloadable OpenAPI 3.1 spec. ([source](https://docs.gable.ai/api-reference/introduction))
- dbt Integration: No — No dbt integration is documented; lineage comes from Java and TypeScript service code. ([source](https://docs.gable.ai/docs/onboarding/coverage-and-completeness))
- Python SDK: Unknown — The gable package on PyPI is described as a CLI and SDK, but the docs only document CLI commands, not a Python library API. ([source](https://docs.gable.ai/docs/cli/introduction))
- Webhooks: Yes — Webhook endpoints with custom headers and a secret receive action events such as data asset changes. ([source](https://docs.gable.ai/api-reference/webhook/create-a-webhook-endpoint))

### Data Context Layer

- MCP Server: No — No MCP server is documented. ([source](https://docs.gable.ai/api-reference/introduction))
- Knowledge Graph: No — Lineage is a graph of components, payloads and fields, but there is no property graph of users, policies and terms. ([source](https://docs.gable.ai/docs/tracing/end-to-end-lineage))
- Semantic Models: No — No semantic model generation is documented. ([source](https://docs.gable.ai/intro))
- Verified Queries: No — Gable has no AI assistant or agent that answers business questions, so there are no curated queries for one to use. ([source](https://docs.gable.ai/docs/onboarding/services-and-data-components))
- Semantic Layer Interop: No — Gable models services, data stores and field mappings; no metric or semantic model import or export is documented. ([source](https://docs.gable.ai/docs/onboarding/services-and-data-components))
- Unstructured Data Catalog: No — Components are scanned services and data stores (tables, Kafka streams, files) tracked for field lineage; documents and vector indexes are not cataloged. ([source](https://docs.gable.ai/docs/onboarding/services-and-data-components))

### Security

- SSO: Yes — SAML 2.0 SSO with Okta, Google Workspace or any SAML IdP, with accounts created on first login. ([source](https://docs.gable.ai/docs/settings/sso))
- RBAC: Yes — Viewer, Editor and Admin roles control read, write and delete across UI and API. ([source](https://docs.gable.ai/docs/availability_security_compliance/role_based_access_control))
- Audit Log: Unknown — Gable keeps CloudTrail logs of its own AWS infrastructure, but no in-app log of user logins or settings changes is documented. ([source](https://docs.gable.ai/docs/availability_security_compliance/security_and_compliance))
- Compliance Certifications: Yes — Gable completed its latest annual SOC 2 Type 2 audit in June 2025; the report is available on request at trust.gable.ai. ([source](https://docs.gable.ai/docs/availability_security_compliance/security_and_compliance))
- SCIM Provisioning: No — Users are created on first SSO login and must be removed in the IdP; no SCIM sync is documented. ([source](https://docs.gable.ai/docs/settings/user_management))

### Deployment

- SaaS Deployment: Yes — Each customer gets its own Gable-managed AWS account, running on serverless AWS services. ([source](https://docs.gable.ai/docs/availability_security_compliance/security_and_compliance))
- Self-Hosted Deployment: No — The platform runs only in Gable-managed AWS accounts; only the CLI scanner runs on customer machines. ([source](https://docs.gable.ai/docs/availability_security_compliance/high_availability_disaster_recovery))
- Hybrid Deployment: Yes — The CLI or Docker image scans code on the customer's CI/CD runner and sends only code-level metadata to the Gable platform. ([source](https://docs.gable.ai/docs/availability_security_compliance/data_access))
- Data Residency: Unknown — Each customer gets a primary and secondary AWS region, but the docs don't say whether customers can choose them. ([source](https://docs.gable.ai/docs/availability_security_compliance/high_availability_disaster_recovery))

### Open Source

- Open Source: No — The platform is proprietary; Gable's GitHub hosts the MIT-licensed Recap schema library, not the product. ([source](https://github.com/gabledata))

## Integrations

9 integrations, each sourced from the vendor docs.

- Authentication: Google Workspace, Okta, SAML
- Code Repositories: GitHub
- Data Orchestration: GitHub Workflows
- LLM Providers: Anthropic, OpenAI
- Notifications: Slack, Webhook

## Alternatives

- [Foundational](https://toolsfordata.com/tools/foundational/): The closest match: lineage from source code, with impact shown in pull requests.
- [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 column-level lineage from the warehouse to dashboards, inside a full catalog.
- [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.
