Anthropic and Snowflake expanded their partnership with a $200 million deal to embed Claude more deeply into Snowflake’s governed data platform and accelerate production deployment of enterprise AI agents across major cloud environments. (Source: Image by RR)

Use Cases Include Regulated Industries Like Finance, Healthcare and Life Sciences

Anthropic and Snowflake announced a major expansion of their strategic partnership, signing a multi-year deal valued at $200 million and launching a joint go-to-market initiative aimed at deploying AI agents inside large enterprises. The agreement makes Anthropic’s Claude models available across the Snowflake platform for 12,600+ customers, spanning deployments through Amazon Bedrock, Google Cloud Vertex AI and Microsoft Azure. The pitch is straightforward: bring frontier-model reasoning directly to governed enterprise data where it already lives, with security controls designed for regulated industries. Both companies frame the move as a shift from experimentation to production-grade deployments.

The partnership, as noted in anthropic.com, is also operationally embedded inside Snowflake itself. Snowflake says it uses Claude Code to boost engineering productivity, and it has built a Claude-powered GTM AI Assistant on Snowflake Intelligence to help sales teams centralize data, ask questions in natural language, and accelerate deal cycles. On the customer side, Snowflake reports that thousands of clients already route trillions of Claude tokens per month through Snowflake Cortex AI, signaling sizable existing usage. The expanded collaboration is meant to deepen that footprint by standardizing how enterprises build and run agentic workflows on top of their Snowflake environments.

Technically, the “next phase” emphasizes agents that can handle complex, multi-step analysis by pairing Claude’s reasoning with Snowflake’s governed data and AI tooling. Snowflake claims 90%+ accuracy on complex text-to-SQL tasks based on internal benchmarks, enabling business users to ask questions in plain English while the agent determines what data is needed, retrieves it across Snowflake, and returns an answer. The offering spans structured and unstructured data, with support for multimodal workloads via Cortex AI Functions (including querying text, images, and audio through SQL). Governance and observability are positioned as a differentiator, with Snowflake’s Horizon Catalog meant to provide responsible AI controls that help regulated teams confidently move from pilots to production.

Early customer examples highlight how the combined stack is being used in production settings. Simon Data is cited as using Claude on Snowflake to uncover patterns while maintaining strict governance, while Intercom uses Claude through Snowflake Cortex AI to power its Fin AI Agent, claiming improved automation and more efficient enterprise engagements. The companies also outline a wealth-management scenario where a Claude-powered agent synthesizes holdings, market data, and compliance rules to generate portfolio recommendations without leaving Snowflake’s security perimeter. Customers can start via a quickstart guide, and Anthropic reiterates a distribution advantage: Claude is available across all three major cloud platforms—Bedrock, Vertex AI, and Azure.

read more at anthropic.com