Snowflake Inc. announced a series of enterprise AI innovations, led by the general availability of Semantic View Autopilot, aimed at making AI agents more trustworthy, governed, and scalable in production environments.

Semantic View Autopilot automates the creation, optimization, and governance of semantic views, giving AI agents and analytics tools a shared, consistent understanding of business metrics. Snowflake said the service can cut semantic model creation from days to minutes and reduce AI hallucinations by ensuring agents operate on the same governed business logic across tools such as dbt, Looker, Sigma, and ThoughtSpot.

The company also introduced expanded capabilities across machine learning and agent observability. Snowflake Notebooks, now generally available, integrates with Cortex Code to let users build and deploy end-to-end ML pipelines using natural language prompts. New features such as Online Feature Store and Online Model Inference support real-time model serving at scale, while Cortex Agent Evaluations (generally available soon) enable enterprises to audit, measure, and validate AI agent behavior before production deployment.

Snowflake said the updates strengthen its AI Data Cloud by unifying trust, governance, execution, and cost control, helping enterprises deploy AI systems that are reliable, transparent, and economically sustainable at scale.

Source: Business Wire