Immuta CEO on Defending Access, Ownership in Data WarehousesMatthew Carroll on Why Heterogeneous Cloud, App Setups Make Data Defense Difficult
The shifting of information to data warehouses such as Snowflake and Databricks has created oversight challenges around access and ownership, said Immuta CEO Matthew Carroll.
Organization should be able to scan and analyze where all their cloud data lives, identify and fix any vulnerabilities or abnormalities, and lock down third-party access to information in data warehouses, Carroll said. But the number of SaaS tools, apps, cloud storage and public cloud options makes guarding data difficult since the same organization might use AWS, Azure and Google Cloud in different parts of the world.
"The broadness is the challenge," Carroll said. "Where do you start to protect? Where do you start scanning for risk? We're trying to work with Global 2000 customers first because they have the broadness problem. You can't solve everything all at once. You can't protect everything. Over the next few years we may be able to, but not initially."
In this video interview with Information Security Media Group, Carroll also discussed:
- How Immuta's technology aligns with what data warehouse companies do;
- Key differences between scanning AI workloads and conventional workloads;
- Why smaller language models have less short-term risk than larger ones.
Carroll's mission is to make the future of data secure. He is known for building and securing scalable data systems, as well as his service and innovation within the U.S. federal government. Before co-founding Immuta in 2015, Carroll served as an intelligence officer in the U.S. Army, including tours in Iraq and Afghanistan. After his military service, he served as CTO of CSC's Defense Intelligence Group, where he led data fusion and analytics programs and advised the U.S. government on data management and analytics issues. While supporting the U.S. intelligence community, he analyzed some of the world’s most complex data sets and solved challenging data management, analytics and intelligence problems in the field as a forward-deployed engineer.