Security Enables Modern Data Adoption
Zachary Friedman, associate director of product management at Immuta, on security in the modern data stack
The strategic point is that security has shifted from a blocker on warehouse adoption to a software layer that makes warehouse adoption possible. Modern data teams already want Snowflake, Databricks, BigQuery, and similar platforms because that is where analytics work now happens. The problem is no longer whether companies want the stack, but whether they can let hundreds or thousands of employees query one shared copy of sensitive data without duplicating tables or hand managing grants.
-
Immuta sits in the gap between identity systems like Okta and data platforms like Snowflake or Databricks. A company connects users and tables, then writes one business level rule that can be enforced across multiple platforms. That matters most when access decisions are decentralized across many teams and data stores.
-
The modern data stack made this urgent by pushing more self serve analysis into the warehouse. As more non specialists run interactive queries, the old workaround of making separate copies of data for different groups breaks down, because it adds compliance risk, operational mess, and constant change management.
-
This is different from adjacent vendors like BigID. BigID starts with finding and classifying sensitive data across many systems for privacy and compliance workflows. Immuta starts closer to live access control inside analytical platforms, then layers discovery, detection, and audit on top so one query can return different rows or masked fields for different users.
Over the next few years, the winning products in this layer will turn secure access into background infrastructure. As warehouses become the default place where companies store, join, and act on data, access control will be expected to work across clouds, catalogs, and AI workflows with almost no user friction. That pushes the category from niche governance software toward core data platform plumbing.