Snowflake rolls out updates to Intelligence, Cortex Code products

Snowflake has announced updates to its Snowflake Intelligence and Cortex Code products.

Snowflake Intelligence now serves as a personal work agent for business users that adapts over time by learning individual preferences and workflows to deliver more relevant results and automate tasks, while also enabling "deep insights" grounded in governed enterprise data.

At the core of this evolution, Skills allow users to describe workflows in natural language — such as preparing presentations, conducting multi-step analysis, or sending follow-ups — and Snowflake Intelligence executes them automatically.

New Model Context Protocol (MCP) connectors allow Snowflake Intelligence to connect directly with enterprise tools like Gmail, Google Calendar, Google Docs, Jira, Salesforce, and Slack so users can operate across the systems they already use.

With deep research, Snowflake Intelligence helps users answer complex questions with fully cited, multi-step reports. It uses an agentic architecture to reason across structured data, unstructured content, and external context, complementing extended thinking's answers with deeper analysis so users can understand not just what's happening, but why and what to do next.

Instead of starting from scratch each time, Snowflake Intelligence now continuously learns from user behaviour to deliver more relevant, personalised responses and automate recurring tasks, while Artifacts allow users to save and share analyses, visualisations, and workflows with each other.

In addition, Cortex Code is expanding as a builder layer for enterprise AI, bringing governed, data-native development across the enterprise data ecosystem so builders can create, orchestrate, and operationalise AI directly within the tools and systems they already use.

Cortex Code now supports even more external data systems including AWS Glue, Databricks, and Postgres, as well as now plugging into other AI systems through the MCP and Agent Communication Protocol (ACP), allowing builders to interface with Cortex Code from their existing AI agents and workflows.

With the new VS Code extension and a Claude Code plugin, builders can access Cortex Code directly in their integrated development environments so they can build and work within their preferred editor or AI coding environment without switching tools.

A new Agent Software Development Kit with support for Python and TypeScript enables teams to integrate Cortex Code's capabilities directly into their own apps and workflows — moving Cortex Code from a standalone tool to a platform other systems can build on.

With Cloud Agents, users can now run code and execute workflows directly in their browser, extending the capabilities of Cortex Code beyond the CLI into a fully managed cloud environment, with no local setup required.

New enhancements including Plan Mode lets users preview and approve workflows before execution, while Snap & Ask enables direct interaction with data artifacts like charts and tables to improve accuracy and give teams more control over how work gets done.

“AI is changing how every company operates, and the platforms that win will make it easy to put AI into practice with the right data and guardrails,” said Baris Gultekin, VP of AI at Snowflake.

“Snowflake gives customers one place to bring their data together, connect the systems they rely on, and turn AI into something that actually helps teams get work done.”