Glean and Snowflake Partner to Make Enterprise Data Instantly Usable Through AI Agents

Enterprise AI adoption often stalls at the point where strategy meets reality: employees have questions, but the answers are in data they can’t access. While structured data lives in warehouses like Snowflake, and unstructured knowledge is scattered across communication and productivity tools, surfacing relevant insights has remained a manual and siloed process.

A new partnership between Glean and Snowflake aims to reduce that friction. On Tuesday, Glean announced a direct integration with Snowflake Cortex Analyst, allowing business users and AI agents to query structured data stored in Snowflake using natural language or SQL, directly from within the Glean platform.

The integration gives employees a single interface to search across both structured and unstructured data without needing to understand database syntax. Glean’s core product already indexes information across platforms like Gmail, Google Drive, Jira, Slack, and Confluence. With Snowflake Cortex Analyst now embedded, users can extend their queries to include real-time data housed in Snowflake such as financial metrics, operational dashboards, or customer records.

“Every employee should be able to ask a question and get an answer—no technical skills required, no time lost waiting,” said Tamar Yehoshua, President of Product and Technology at Glean. “With our support for Snowflake Cortex Analyst, we’re closing the gap between data and decisions.”

For many teams, accessing Snowflake data has typically required data analysts or business intelligence teams to run custom queries. The addition of natural language support removes that barrier. Glean users can now ask questions like “What were this month’s top support issues by volume?” or “How did sales pipeline growth compare across regions last quarter?” and receive structured responses based on Snowflake data, augmented by relevant files or documents from other systems.

Baris Gultekin, Head of AI at Snowflake, emphasized the flexibility this brings to joint customers. “We’re making it easier for customers to harness the power of their structured data,” he said, “and providing more choice in how and where they use Cortex Analyst to unlock business insights.”

Beyond natural language access, the partnership also extends to Glean’s growing platform of AI agents—prebuilt and customizable assistants that help automate routine tasks and retrieve insights on demand. With the new integration, Glean agents can include Snowflake queries as part of broader workflows, such as generating sales reports, analyzing churn trends, or preparing quarterly business reviews.

These agentic workflows benefit from Glean’s internal “system of context,” a proprietary indexing engine that maps relationships between users, content, tools, and permissions across an organization. This allows both users and agents to perform actions in a way that is consistent with enterprise access rules and contextual relevance.

Security and governance remain core to the implementation. Glean has built its platform to enforce permission boundaries at every step, ensuring that neither users nor agents can access data they aren’t authorized to see. The company’s security framework includes protections against common AI threats, such as prompt injection, and complies with enterprise-grade standards around data residency and access auditing.

As part of the release, Glean has also expanded its developer ecosystem. Customers can build their own agents using Glean’s natural language agent builder or through its API, which supports TypeScript, Java, Python, and Go. Developers can define the models used at each step of an agent’s process, set permissions, and tailor workflows to specific roles or departments.

Glean has introduced a “model hub” that gives customers access to a range of LLMs through platforms like Amazon Bedrock, Google Vertex AI, and Azure OpenAI. Customers can mix and match models based on task needs, choosing factual response models for legal agents or more generative ones for marketing. The platform includes three “temperature” settings—factual, balanced, and creative—to help align tone and reliability with intended use cases.

The company reports that customers are on pace to complete over 100 million agent actions this year. These include both individual productivity tasks and department-level automations. For instance, Booking.com created a Glean agent to extract user insights from product research reports, which were previously difficult to access. A global social platform used Glean to build a threat modeling agent that accelerated its security workflows by 97%, cutting time and cost significantly.

To support broader enterprise deployment, Glean has also introduced administrative tools that centralize agent governance, audit usage, and manage access across teams. Organizations can curate a library of approved agents and customize them as business needs evolve.

The Snowflake partnership builds on this strategy by bringing structured data queries into the same environment. For teams that rely heavily on Snowflake for operational metrics or analytics, the integration creates a pathway to embed those insights into everyday tools and decision-making processes without having to train the entire workforce in SQL or BI platforms.

“The enterprise brings a lot of complexity—fragmented systems, inconsistent permissions,” said T.R. Vishwanath, Glean’s co-founder and CTO. “The first step is to make everyone powerful.”