Snowflake is significantly expanding its artificial intelligence capabilities with enhanced offerings for Snowflake Intelligence and Cortex Code, aiming to embed AI development and deployment within its ecosystem. This strategic move caters to a broad spectrum of users, from generalized business professionals to specialized development and technical teams, through a dual-pronged approach.

Snowflake Intelligence is meticulously designed for non-technical business users, offering a intuitive platform to integrate and execute AI tasks directly within existing business workflows. This platform empowers users to articulate desired actions in natural language, which the underlying large language model (LLM) then processes to perform various tasks. Envisioned uses include sophisticated presentation preparation, multi-step data analyses, and automated follow-up communications. The system draws data from an organization’s internal digital assets, encompassing both structured and unstructured data, and connects to external sources via a variety of protocols and pre-built connectors. To ensure data integrity and compliance, user queries and subsequent workflows are rigorously governed by access permissions and organizational protocols, mitigating risks of data loss and non-compliance. New interfaces leveraging the Model Context Protocol (MCP) are now available, with integrations extending to prominent business tools like the Google business suite, Jira, and Salesforce (including Slack). Furthermore, an iOS application for Snowflake Intelligence is slated for public preview soon, enhancing mobile accessibility. The platform is designed to become more personalized over time, learning from user behaviors, and allowing users to save and share workflows for reusability. Enhanced context windows mean users will experience a more fluid interaction, reducing the need to repeat extensive contextual prompts.

The evolution of Snowflake Intelligence is a direct result of feedback garnered from Project SnowWork, a research initiative launched recently to showcase the platform and collect user preferences for AI features. This user-centric development underscores Snowflake’s commitment to delivering highly relevant and effective AI solutions.

Conversely, Cortex Code is tailored for enterprise software development teams, providing a robust coding and orchestration layer where AI algorithms can efficiently handle lower-level tasks. This platform has expanded its integration capabilities with external data sources, now supporting AWS Glue, Databricks, and Postgres. Cortex Code also facilitates connections to other language models through MCP and the Agent Communication Protocol (ACP), the latter being a commerce-driven protocol. Developers using VS Code can anticipate a Cortex Code extension, currently in private preview, and a dedicated Snowflake plugin for Claude Code is actively under development. The Agent Software Development Kit (SDK) for Python and TypeScript is already available, enabling teams to seamlessly embed Cortex Code functions into their proprietary applications. Cloud Agents, also in private preview, will soon be accessible within Snowsight, Snowflake’s browser-based interface. A key feature, Plan Mode, allows users to preview and approve workflows before AI execution, and Snowflake is developing a mechanism for end-users to gain detailed insight into the LLM’s longer research processes to verify the veracity of its operations.

Snowflake's AI offerings have seen significant adoption, with over 9,100 customers utilizing its AI products weekly. Remarkably, more than half of its customer base has adopted Snowflake Intelligence and Cortex Code within six months of their launch. This dual strategy of addressing both mainstream business users and specialized software development teams reinforces Snowflake's core technical market while broadening its platform's appeal across general business functions. The introduction of new software connectors, a mobile application, and browser-based options is set to expand the user base significantly. Moreover, enhanced support for existing enterprise systems will further solidify its appeal among organizations with deeply embedded workflows and software platforms. Sameer Vuyyuru, Chief AI and Product Officer at Capita, lauded Snowflake’s ability to deploy AI securely with appropriate governance, which is critical for highly regulated, citizen-facing services demanding performance, compliance, and trust.