A data scientist for everyone.
Our most-used agent at Notion brings together our Snowflake data and the business context needed to interpret it. Anyone on our team can ask a question in plain language and get a grounded answer in seconds.
Wait 2 weeks for that report. Or 2 minutes.
Connect Notion agents and Snowflake data for answers you can take to your next meeting, turn into a report, or share with your team.
Uh-oh! It looks like your ad blocker is preventing the video from playing.
Turn findings into a report.
Generate a Notion report your team can review and share.
Ask on the go.
Put your data agent in your pocket with the Notion mobile app.
How it works.
Notion for company context, Snowflake as the governed data foundation.
Data agents in action.
More than 60% of Notinos across GTM, Product, and Ops use our data agent daily. Easier data access helps teams get answers, while our data team handles 5x fewer requests, freeing their time for higher-impact work.
Notion’s data agents make a lot of sense for Qonto. Our data is governed in Snowflake but our people are in Notion, so this integration gives us a path to more reliable reporting and faster time-to-insight across teams.
Our customers have already done the hard work of building a trusted, governed data foundation in Snowflake. By integrating with Notion, we're removing the friction between that data and everyday action.
Build with enterprise confidence.
Grounded in your business
Use the metric definitions, table context, company knowledge, and query guardrails set by your teams.
Respects permissions
Admins control who uses the agent. Users sign in with their own Snowflake credentials, so existing permissions apply.
Sources you can check
Review the data, definitions, and company context behind an answer before taking action on it.
Admin controls and analytics
Workspace admins can monitor agent usage and Notion credit consumption. Every agent run is logged with full audit trails showing what triggered it, what it did, and why.
Build a data agent for your team.
Request early access to discuss your use case and explore whether the program is a fit for your organization.
FAQs
How does a data agent work with Snowflake and Notion?
A data agent brings together governed Snowflake data, relevant company knowledge, and clear instructions. When someone asks a question, it uses approved definitions and source guidance to investigate, then returns an answer with supporting sources. Data access is limited by the user’s Snowflake role and permissions.
Notion gives teams a place to ask questions, review results, and put them to use in their work.
For more detail, read How we built a personal data scientist for every Notion employee.
How are permissions & access governed?
A data agent can be implemented either as Custom Agent connected to Snowflake MCP, or invoked as a skill in Notion Agent. In both instances, each user is required to connect to Snowflake using their own credentials. Notion controls workspace access and available connections, while Snowflake enforces the approved role and data permissions for the person that's asking the question.
Your data team defines the metrics, sources, filters, and query guardrails. If the data agent runs into missing definitions, failed checks, or if the user has insufficient Snowflake access, the agent surfaces those details instead of guessing the answer.
How can I learn more about building a data agent?
For a look at how we built our data agent with Notion AI and Snowflake, read How we built a personal data scientist for every Notion employee. If you’d like help building one for your team, request early access.
Who is the Early Access program for?
This is a guided program for teams with a recurring business question, trusted Snowflake data, and the people and time to build and test a first workflow together. A strong fit has a business champion, Notion and Snowflake owners, a data owner, a small committed pilot group, and a path through security review. Teams evaluating either platform can still express interest, and our team will review readiness and fit.
What happens after I request access?
Our team will review your use case and readiness. If accepted, we’ll guide your team through discovery, setup, and validation of one initial workflow. Because spots are limited, some qualified teams may be waitlisted.