Immuta announced that several of its retrieval-augmented generation-based generative AI solutions across various cloud platforms have received new data governance and audit features, SiliconAngle reports.
RAG applications enhance large language models by integrating external knowledge sources to improve content accuracy. Immuta's multilayer architecture secures, monitors, and audits sensitive data accessed by these AI applications, addressing the challenge of managing corporate data risk amid increasing AI usage.
The solution targets three defense layers: storage, data, and prompt. Immuta emphasizes securing the first two layers, partnering with Amazon Web Services to create a native S3 storage integration that enforces fine-grained access control. The data layer transforms unstructured data for model training and RAG use, with capabilities for discovering, classifying, and controlling RAG indexes.
Immuta's new features allow data teams to manage access with multilayered policies, maintain accurate metadata inventories, and control RAG-based applications by creating natural language policies. The platform supports Snowflake, Databricks, and provides comprehensive monitoring and auditing across all supported platforms.
“Data teams are now able to leverage the significant investments they have made in their cloud data platforms and rapidly extend this work to their AI application workloads," according to Immuta Chief Product Officer Mo Plassing.
