Collibra, a leader in unified data and artificial intelligence governance, announces a major enhancement to its existing integration partnership with Snowflake, during Snowflake’s yearly user event, Snowflake Summit 26. Building on the company’s announcement of its recently launched AI Command Center, which is a unified control platform for enterprise-wide AI compliance, the collaboration between Snowflake and Collibra enables a more advanced integration of Collibra’s governance mechanics within the Snowflake AI Data Cloud, providing an extremely secure environment for companies to transition from pilot AI initiatives into production.
With agentic AI models executing autonomous functions and making business critical decisions, their outputs have implications for the company’s revenue growth and compliance with regulations, as well as its reputation with consumers. The time lag between traditional risk assessments and ad hoc model registration and sign-off cannot match the pace at which autonomous solutions operate.
The expanded alliance ensures that Collibra’s structured enterprise taxonomy encompassing critical data definitions, quality scores, ownership parameters, and business summaries natively populates the Snowflake Horizon Catalog, Cortex Analyst, and Cortex Agents. Grounding natural language models and automated agents in predefined enterprise standards improves output precision and mitigates structural compliance risks.
Also Read: IBM Unveils Managed Red Hat AI Inference and OpenShift Virtualization Services to Scale Enterprise Hybrid Cloud
“Data and AI are only as valuable as the context around them, and our expanded collaboration with Collibra is about delivering that context and the control to act on it, seamlessly to our customers,” said Amy Kodl, SVP of WW Partner and Alliances at Snowflake. “By enabling bi-directional metadata synchronization with Snowflake and grounding Cortex Analyst and Cortex Agents in business semantics, we’re helping organizations accelerate their path to data driven insights and trusted AI at production scale.”
“The future of data lies in making context and governance an inherent part of the AI fabric,” said Chandra Papudesu, VP, Product Management, Integrations & Lineage at Collibra. “This strengthened partnership with Snowflake, and the bidirectional integration, gives our joint customers the foundation to scale agentic AI from experimentation to production with confidence.”
Bi-Directional Synchronization Across Snowflake Horizon Catalog and Cortex AI
The centerpiece of this expanded architectural capability is a native, bi-directional synchronization channel. Under this design, Collibra’s policy libraries, taxonomies, and descriptions flow directly into Snowflake, while Snowflake’s technical lineage and structural metadata simultaneously feed back into Collibra. This continuous loop aligns business intent with the underlying data architecture.
- Governance and Policy Enrichment (Collibra to Snowflake): Enterprise criteria established within Collibra integrate directly into Snowflake, supplying immediate policy parameters and secure data guardrails for analytical teams across the enterprise.
- Semantic Enrichment for AI (Collibra to Snowflake): Semantic models created within Collibra conform to the OSI semantic model standard. OSI is vendor-agnostic and supported by Collibra. They may be exported to Snowflake in either the OSI standard or in the form of Snowflake Semantic Views.
- Automatic Data Traceability (Snowflake to Collibra): The metadata information and lineage details are automatically extracted from Snowflake and imported back to Collibra. This ensures that the latest data concerning your data environment is stored, and you have an up-to-date system of engagement for Cortex Analyst and Cortex Agent users.
- Capacity-based Access Management Controls: The security management team can automate the assignment, audit, and revocation of fine-grained access controls to both Collibra and Snowflake.
Collibra AI Command Center: The Control Plane for Snowflake-Based Agentic AI
With the help of the Collibra AI Command Center, the complete transparency is achieved in all phases of data, model weights, automated decisions, and finally business impact. Through the Collibra MCP Server on the platform, real-time data certifications, policies, and quality scores are available to enterprise agents prior to any action being taken.
In order to fulfill the requirements of model risk management, the solution enables organizations to move forward with their automated AI capabilities.
