Snowflake expands its technical and mainstream AI platforms

Snowflake is expanding its Snowflake Intelligence and Cortex Code offerings in an effort to support users deploying and developing artificial intelligence across the Snowflake portfolio. Snowflake Intelligence is positioned as a tool for general business users, while Cortex Code is aimed at developers and technical teams.

According to the company, both platforms are receiving additional features, including expanded integrations with third-party software, new automation capabilities, and simpler, web-based methods for building agentic AI workflows.

Snowflake Intelligence for business users

Snowflake Intelligence is designed for non-technical staff and focuses on executing tasks within existing business workflows. Users can describe in natural language what they want to achieve, and the large language model (LLM) carries out the requested actions.

Use cases highlighted by Snowflake include preparing presentations, running multi-step analyses, and sending follow-up messages. The platform can draw on an organisation’s internal and linked digital assets, including both structured and unstructured data, as well as external sources connected via various protocols and pre-built connectors.

User queries and resulting workflows are constrained by access permissions and organisational governance policies, with the aim of preventing data loss and non-compliance.

New interfaces using the Model Context Protocol (MCP) are now available. Snowflake Intelligence can interface with systems including the Google business suite, Jira, and Salesforce (including Slack). An iOS app for Snowflake Intelligence is also in development and is expected to enter public preview “soon.”

Snowflake says the Intelligence platform becomes more personalised over time by learning from user behaviour. Users can save and share workflows for reuse. Longer context windows are intended to reduce the need for repeatedly providing lengthy prompts to achieve desired results.

The updates follow feedback from Project SnowWork, a research initiative launched to showcase the platform and gather user preferences on desired AI features.

Cortex Code for development teams

Cortex Code is built for enterprise software development teams, where AI can assist with lower-level development tasks. Snowflake describes it as a coding and orchestration “layer” with expanded integration options for external data sources, now supporting AWS Glue, Databricks, and Postgres.

Cortex Code can also connect to other language models via MCP and the Agent Communication Protocol (ACP), a commerce-driven protocol that emerged around the same time as Anthropic’s MCP.

A Visual Studio Code extension for Cortex Code is currently in private preview, and a Snowflake plugin for Claude Code is under development. Snowflake’s Agent Software Development Kit (SDK) for Python and TypeScript is available, enabling teams to embed Cortex Code functions into their own applications.

Cloud Agents, also in private preview, are set to appear in Snowsight, Snowflake’s browser-based interface. Plan Mode allows users to preview and approve workflows before AI execution. The company is also working on a feature that will let end-users view details of longer research processes undertaken by the LLM to verify its outputs.

Enterprise adoption and governance

Snowflake reports that more than 9,100 customers use its AI products weekly. Since their launch six months ago, more than half of Snowflake’s customers are using Snowflake Intelligence and Cortex Code.

The company’s dual approach—serving both mainstream business users and software development teams—extends its core technical market while broadening adoption across general business functions. Expanded connectors, mobile access, and browser-based tools are intended to increase usability and appeal among enterprises with established workflows and software ecosystems.

“Snowflake helps us deploy AI securely and with the right governance across highly regulated, citizen-facing services where performance, compliance and trust are critical,” said Sameer Vuyyuru, chief AI and product officer at Capita.