Defining Snowflake’s New Enterprise AI Direction
Snowflake’s acquisition of Natoma and its expanded AWS partnership represent a shift in enterprise AI from model experimentation toward secure, action-oriented agents tightly integrated with governed data and business applications. Instead of focusing only on answering questions, Snowflake is building what CEO Sridhar Ramaswamy calls an “agentic control plane,” where AI agents can take actions inside email, chat, calendars, ticketing systems, and cloud tools while staying within enterprise security and compliance. This approach shows how enterprise AI priorities are moving from raw model performance to safe deployment, identity-aware access, and detailed audit trails across complex technology stacks. By combining a cloud data platform with an agent governance layer and deeper cloud infrastructure integration, Snowflake aims to become the core environment where enterprises design, monitor, and scale AI agents in production.
Natoma Acquisition: Governance for Rogue AI Agents
Natoma sits at the center of Snowflake’s evolving acquisition strategy, aimed at making AI agents reliable enough for real enterprise work. Natoma builds a gateway for Model Context Protocol (MCP) servers, allowing AI agents to call tools like Slack, Google Docs, Jira, and internal APIs while enforcing identity checks, fine-grained permissions, and audit controls for every tool call. Users can send emails, summarize Slack conversations, check calendars, and open Jira tickets from products such as Snowflake Intelligence and Cortex Code, with all actions logged and governed. Natoma’s founders describe their platform as a “secure connectivity, identity, and governance layer” that extends Snowflake experiences safely into existing applications. This directly addresses concerns about rogue or shadow agents acting outside policy, aligning the Snowflake acquisition strategy with rising demand for AI governance and controlled access rather than unrestricted automation.
Sixth Acquisition Since 2025: A Data-to-Agent Platform Play
Natoma is Snowflake’s sixth acquisition announcement since June 2025, marking a clear acceleration in its push from data warehousing toward a full cloud data platform for AI. Earlier deals targeted core database and analytics capabilities, such as PostgreSQL provider Crunchy Data and database migration specialist Datometry, along with data discovery outfit Select Star. More recent transactions have focused on AI-native infrastructure, including Observe, an AI observability platform, and TensorStax, which plans AI-powered data pipeline planning. Natoma extends this acquisition thread into the realm of agent governance and identity. Together, these purchases show Snowflake acquisition strategy evolving from classic analytics toward an integrated stack for data, pipelines, observability, and now agentic control. The company is positioning itself not only as the place where enterprise data lives, but also where AI agents are designed, monitored, and constrained as they act on that data.
AWS Partnership Expansion and AI Infrastructure Demand
Alongside the Natoma integration, Snowflake announced an expanded, five-year AWS partnership that includes a USD 6 billion (approx. RM27.6 billion) infrastructure commitment centered on Graviton-powered compute and AI services. This AWS partnership expansion shows how enterprise AI priorities now depend on tight alignment between cloud infrastructure and data platforms. Snowflake and AWS plan deeper integrations around generative and agentic AI, broader AWS Marketplace sales motions, and joint programs to help customers migrate workloads. According to AWS CEO Matt Garman, “Enterprises are rapidly moving from experimenting with AI to putting intelligent agents to work that drive real business outcomes.” By keeping AI models close to governed data inside Snowflake, running on AWS infrastructure, both firms aim to reduce data movement, improve security, and support the heavy compute needs of large agent deployments across industries and workloads.
Competitive Positioning and Future Enterprise AI Priorities
Snowflake’s dual focus on Natoma integration and AWS alignment signals a new phase for enterprise AI priorities: governed agents over general-purpose chatbots, and platform depth over one-off tools. Internal research cited by Snowflake shows that 96% of organizations still face major obstacles when scaling AI, especially around governance, permissions, and integration with existing workflows. By building an “agentic enterprise” vision on top of its cloud data platform, Snowflake is betting that customers will prefer AI systems grounded in their own data, governed by their own policies, and wired into their existing SaaS and infrastructure. This competitive positioning challenges AI-first vendors that sit outside core data platforms, as well as infrastructure players lacking native governance. If Snowflake can prove that its agentic control plane reduces risk while improving productivity, it may become a default choice for enterprises standardizing their AI stack.