Native Snowflake, shipped

Snowflake is one of the most common data warehouses in our customer base, and the integration we've shipped today removes the awkward setup that used to be required. Argus now connects natively — no scanning infrastructure to host, no IP allowlists to maintain.

What's new

The previous Snowflake support worked but required customers to run a scanning agent in their own infrastructure, which created operational overhead nobody wanted. The native integration replaces that with direct connectivity through Snowflake's external function and stored procedure interfaces, authenticated through OAuth or key-pair credentials managed in our platform.

The result: setup is three steps in our admin interface, scans run on Snowflake's compute (within limits you configure), and there's nothing to deploy or maintain on the customer side.

What it covers

Schema-level discovery across all databases and schemas Argus has been granted access to. Column-level classification across structured and semi-structured data — including JSON columns, where we sample values to detect sensitive content rather than relying on column names. Access pattern monitoring that surfaces unusual queries against sensitive columns.

For customers using Snowflake's row-level access controls and dynamic data masking, Argus can also flag policy gaps where sensitive columns are exposed to roles that shouldn't have them. This is one of the more useful capabilities for governance teams trying to verify their masking strategy is doing what they think it's doing.

Migrating from the legacy connector

If you're a current customer running the legacy connector, you can switch at any time from the integrations page. The migration preserves all of your existing classifications, policies, and historical scan data — you don't lose continuity in your audit logs or data lineage. We'll keep the legacy connector supported through the end of the year for customers who need time to transition.