A woman focused on her laptop in a modern office setting, with two large monitors visible and a cup of coffee on the desk.

Snowflake Cortex Code sparks AI-powered development shift

Snowflake announced major updates to its AI coding assistant Cortex Code in March 2026, introducing multi-agent collaboration teams and specialized domain skills to boost developer productivity. The cloud data company made the tool generally available within its Snowsight interface and added Windows support, enabling users to generate 90% of their code through AI while completing tasks in hours instead of days.

The enhancements position Snowflake to compete more aggressively in the enterprise AI coding market, where developers increasingly rely on intelligent assistants to accelerate workflow. The company emphasized that these capabilities are available with “zero setup” to all Snowflake users, following the platform’s existing consumption-based pricing model rather than requiring separate licenses.


The centerpiece of the update is Agent Teams, a multi-agent orchestration system that fundamentally changes how complex coding tasks are handled. According to Snowflake’s announcement, this feature enables a primary agent to delegate work to specialized sub-agents that operate in parallel, each taking on distinct roles such as research, coding, and testing while collaborating through a shared task list.


Beyond team collaboration, Cortex Code now includes pre-packaged “Skills” that provide domain-specific expertise. The newly introduced skills cover cost intelligence, machine learning, Streamlit development, and Cortex AI functions, allowing the assistant to deliver more accurate responses without processing its entire knowledge base for each query.

Productivity Gains and Developer Adoption

A woman focused on her laptop in a modern office setting, with two large monitors visible and a cup of coffee on the desk.

Early adopters are reporting dramatic efficiency improvements. One unnamed customer cited by Snowflake indicated that 90% of their code is now AI-generated, with data engineers submitting five times more pull requests than before implementing the tool. The platform’s integration directly into the Snowsight UI, including Workspaces and Notebooks, provides developers with a context-aware assistant that understands their specific data catalogs, schemas, and semantic models.


The native Windows CLI support, released to open preview in March 2026, extends these capabilities to local development environments. Developers can now integrate Cortex Code into tools like VS Code, Cursor, or terminal applications for building and deploying applications and data pipelines.


Security remains central to the platform’s design. Cortex Code operates entirely within users’ secure Snowflake environments, leveraging the Horizon Catalog to ensure AI suggestions automatically adhere to existing data access and governance controls. According to Snowflake, this integration maintains enterprise-grade security while the AI agent assists with everything from generating SQL and Python code to fixing failed queries with single-click solutions.


The updates target a broad user base, from “seasoned engineers to less technical builders,” as Snowflake described in its announcement, signaling the company’s push to democratize AI-powered development across skill levels.

Sources

  • snowflake.com/en/blog