BlogNews & Updates
Snowflake × Fin: Give Fin live data to resolve and route account-specific queries
Our new integration with Snowflake AI Data Cloud grants Fin secure, controlled access to your account data so it can be used to resolve personalized queries.
Today we’re announcing a new Fin integration with Snowflake AI Data Cloud. As organizations adopt AI agents, trusted business context and controlled access to data become essential. With our MCP connector, you can grant Fin controlled, secure access to your Snowflake data, giving it the context it needs to resolve complex, personalized queries.
Here’s what’s new:
- Tailor Fin's answers to every individual customer with live data. With access to your Snowflake account, Fin can answer queries specific to each customer or route them to the right rep.
- Connect without engineering lift. Connect Fin to your tables or Cortex Agents yourself, or let CoCo, Snowflake's AI coding agent, handle the setup for you. Then add your connection details in Fin settings.
- Give Fin secure and controlled access. Fin works within your Snowflake access policies and governance, so no additional configuration is required.
- Connect support to business outcomes. Share in-depth, live conversation data with your Snowflake account and connect the impact of support to downstream business outcomes.
Why we built this
For many companies, the Snowflake AI Data Cloud holds the most complete view of every customer, connecting data across the business: what they’ve ordered, which plan they’re on, how they use your product, and whether their last payment went through.
With this integration, Fin is now able to resolve questions that require account information. When a customer asks where their order is or why their bill changed, they won’t get a general answer or have to wait for a teammate to look it up. Instead, Fin is able to pull the information and respond immediately. When a conversation needs a person, Fin can differentiate between accounts and directly route requests to the right rep, instead of landing them in a shared queue.
In the past, giving an agent access to the full view of customer data meant a custom engineering build, requiring APIs, security reviews, and pipelines to maintain. Most teams don’t have the bandwidth to commit to this, which is why we built an integration with Snowflake that lets Fin seamlessly connect to your existing customer data.
How the integration works
Tailor Fin's answers to every customer with live Snowflake data
Fin connects to the Snowflake-managed MCP server running inside your account and calls the tools your team has configured. These can range from simple searches to using Snowflake Cortex for natural language lookups and analysis.
This allows Fin to securely access account status, order history, product usage, and billing records to resolve customer queries.
The same data also makes handoffs smarter. When a conversation needs a person, Fin can pull the customer’s profile information, such as their account owner and plan tier, and route the request along with a handoff summary. If the customer doesn’t include an identifier, Fin can be configured to ask for the details it needs to find the account.
Connect your data without major engineering lift
Snowflake setup is a one-time process. Scope a role that controls exactly which data Fin can see, set up the MCP server, and configure an OAuth security integration. Once those credentials are in hand, connecting in Fin takes a few seconds – just provide your Snowflake details – and every tool you’ve enabled is available to Fin in conversations.
Give Fin access while staying in control
Your Snowflake administrator controls the integration’s scope and your existing governance and security model are carried over. Fin uses only the tools you enable and the read-only data it has access to stays secure within Snowflake, without copying, moving, or modifying data.
Share live conversation data with Snowflake to connect support to business outcomes
The integration also works in the other direction, passing Fin’s data back to Snowflake. With Snowflake Secure Data Sharing, you can connect your Fin conversations to your Snowflake account so your team can then analyze conversations alongside the rest of your business data and connect support to downstream business outcomes.
How DoorDash uses the Snowflake × Fin integration
DoorDash’s Customer Success team supports about 11,000 active restaurants, each with different needs and a different relationship with the business. Until recently, every merchant request landed in a shared inbox and was handled in the order it arrived. The team had no easy way to triage queries based on account data.
In under two weeks, the team went live with the Snowflake integration. Fin now pulls each merchant’s profile in real time, including order and delivery data, and uses it to route and prioritize every ticket. If a merchant doesn’t include an identifier, Fin asks for the details it needs to find the account.
“Fin and Snowflake gave our Customer Success team something we never had before: real-time context on every business the moment they reach out,” says Jake Kleffman, Associate Manager Strategy & Operations at DoorDash.
Four weeks after launch:
- Average time to close fell by 84%, from 8.7 days to 1.4 days.
- The team spent the same amount of time handling inbound requests, even as DoorDash's merchant base expanded and ticket volume grew by 25%.
“Requests now get to the right person right away, and the quality of care we deliver has improved across the board,” Kleffman says. “We’re excited to bring this to more teams across DoorDash.”
Give Fin the full picture
AI Agents are only as good as the data they have access to. The companies that get the most out of Fin will be the ones that bring their existing data and their support stack together.
Our Snowflake integration helps make that possible, giving Fin the context it needs to resolve and route conversations while your governance stays intact.
See our setup docs for MCP and Secure Data Share or reach out to your Fin Customer Success Manager.