AI Governance & Security Vendor Wiki

5 min read

Alation

Primary category: data-access-governance. Also listed in entitlement-aware-rag.

One-liner — The data catalog that grew into a governance platform: it inventories what data the firm has, where it came from, who owns it and what it means — the metadata layer that determines whether a RAG or agent deployment can be governed at all.

What it does — Alation crawls the firm’s data estate (warehouses, lakes, BI tools, databases) and builds a searchable catalog of tables, columns, dashboards and reports, enriched with lineage (where a figure came from and what transformations it passed through), business glossary terms, ownership, quality signals and usage telemetry — Alation’s original trick was mining query logs to infer which data people actually use and trust. On top of the catalog sit policy definition, stewardship workflow, and a data-products marketplace.

Why an AI-governance wiki cares: the catalog and its policy metadata are the substrate for everything downstream. You cannot restrict what a retrieval pipeline surfaces, prove to an examiner what a model was trained or grounded on, or answer “does this agent have access to MNPI?” without an accurate map of the data estate first. The 2026 Gartner Magic Quadrant for Data and Analytics Governance Platforms made AI model governance a mandatory evaluation criterion — the category’s own recognition that governing the data estate no longer suffices without governing what AI does with it.

Naming / provenance — Founded 2012, Redwood City. Single brand; the market label has drifted from “data catalog” to “data intelligence platform” without a product rename.

Ownership & viability — independent, VC-backed, $340M raised across five rounds. The $123M Series E in November 2022 valued it above $1.7B, led by Thoma Bravo, Sanabil and Costanoa, with Databricks Ventures joining — a strategically significant investor given how much of this market Databricks Unity Catalog and Microsoft Purview are absorbing from below. No round since 2022 and no acquisition as of 2026-08-26. Thoma Bravo on the cap table at a $1.7B mark makes a take-private a natural eventual outcome.

Positioning & differentiators —

  • Versus collibra: the long-standing head-to-head. Collibra is stronger on formal governance process and stewardship workflow; Alation is stronger on discovery, search and analyst adoption. The cliché — Collibra for the governance office, Alation for the people who use data — is roughly right.
  • Versus immuta: a different job entirely. Immuta enforces policy at query time; Alation describes and documents the estate. Firms doing entitlement-aware retrieval seriously often need both — see entitlement-aware-rag.
  • Versus the platform-native catalogs (Databricks Unity Catalog, Microsoft Purview, Snowflake Horizon): the standing threat. Alation’s answer is heterogeneity — it covers estates that span several platforms, which is the normal condition at a firm running Snowflake plus a legacy SQL Server plus vendor market-data feeds.
  • Agentic direction. Forrester named it a Leader in Data Governance Solutions (Q1 2025), citing its agentic AI platform and data-products marketplace. How real the AI-model-governance capability is, versus catalog features relabelled, is an open question below.

Who should choose them / anti-fit — Fits a firm with a heterogeneous data estate and a genuine discovery problem — analysts who cannot find or trust data — that is now trying to point retrieval or agents at internal data. This describes a great many asset managers. Anti-fit: a firm entirely inside one platform (all-Databricks, all-Snowflake, all-Microsoft), where the native catalog is cheaper and better integrated; and any firm whose actual need is enforcement rather than documentation — buy immuta or varonis instead.

Known weaknesses / gotchas — Enterprise price point and a real implementation effort: a catalog is only as good as the stewardship put into it, and half-populated catalogs are the standard failure mode of this category. Squeezed from below by platform-native catalogs. It documents; it does not enforce — a distinction firms routinely blur when they assume buying a catalog gives them access control.

Deployment & data handling — SaaS (Alation Cloud Service) and customer-managed deployment. Alation ingests metadata, not primary data, in normal operation — an important and reassuring distinction — but the specifics of what is transmitted (query logs contain fragments of business content) are unverified and worth diligencing.

Integrations & partnerships — Broad connector library across warehouses, lakes, BI and ETL; Databricks partnership reinforced by the Series E investment. MCP support unverified.

Compliance & FS traction — Unverified here; the customer base is enterprise-wide and financial services is a stated vertical, but no named references were confirmed.

Commercial — Not public. Enterprise subscription, typically scaled by connected sources and user seats.

Open questions

  • What “AI model governance” actually means in Alation’s product versus catalog features relabelled for the 2026 Gartner criterion.
  • Whether query-log ingestion transmits business content, and retention/training terms.
  • MCP or agent-framework support — can an agent query the catalog to resolve permissions and semantics at runtime?
  • Named financial-services customers; certifications held.
  • Any funding or ownership change since the 2022 Series E — Thoma Bravo’s presence makes this worth re-checking.

Sources

History

  • [2026-08-26] Page created via wiki-create + researched same day. Sourced from the 2026-08-25 competitor scan (Tier B, score 7) — surfaced as a peer of Immuta and Collibra; promoted to inclusion because the data-catalog layer is a prerequisite for entitlement-aware retrieval at an asset manager.

Graph View

Backlinks