Immuta Unblocks Cloud Data Migrations
Zachary Friedman, associate director of product management at Immuta, on security in the modern data stack
Immuta matters in these migrations because it turns security from a reason to delay the move into a repeatable software workflow. Large enterprises moving from Hadoop or Teradata to Snowflake or Databricks suddenly have many more analysts querying one shared cloud system, and the old fix of making separate copies of data for each audience breaks under privacy rules, audit needs, and operational sprawl. Immuta sits above the warehouse, connects identity to data, and pushes row, column, and table rules into the underlying platform so teams can keep one governed copy of data while still opening access.
-
The concrete product job is policy translation. A bank or pharma company can define who should see which records and fields using business attributes, then have Immuta automatically convert that into native Snowflake or Databricks controls. That removes manual grants, custom code, and duplicated datasets during migration.
-
This becomes more valuable as companies standardize on cloud warehouses. Snowflake became the centerpiece of the modern data stack, and Databricks scaled into a major enterprise platform on the back of cloud migration. As those systems spread, security moved from an edge concern to a gating requirement for adoption.
-
Immuta is not replacing the warehouse security layer so much as operationalizing it across messy real world environments. Its Databricks Unity Catalog integration uses native row, column, and table controls enforced inside Databricks, while adding centralized policy orchestration that platform teams can use across many datasets and users.
Going forward, the winning data platforms will be the ones that let companies expose more sensitive data to more internal teams without creating more copies or more approval tickets. That pushes Immuta toward becoming the control plane for governed self service analytics, especially as Databricks, Snowflake, and adjacent platforms absorb more AI and multi tenant data workloads.