Overview
The Snowflake Agent Node provides an intelligent agent for querying Snowflake data warehouses. This node enables you to:- Execute SQL queries on Snowflake tables
- Retrieve data from multiple databases and schemas
- Perform complex data analysis and aggregations
- Run synchronous or asynchronous queries
- Configure warehouse, database, schema, and role
- Handle query results and errors
Configuration Parameters
To set up the Snowflake Agent Node, you need to authenticate with your Snowflake account and provide your account URL. Once authenticated, configure the agent to execute SQL operations.Node Configuration
- Account URL: Your Snowflake account URL (e.g., https://myaccount.snowflakecomputing.com)
-
Prompt:
Describe what you want (your intent) in natural language. The agent figures out how to translate that into SQL—you don’t need to write the query yourself.
Total revenue by region in 2024
Top 10 customers by order volume
All products with inventory below 100 units
-
Tools
Select which tools the Snowflake Agent can use:
Tool name Description Execute SQL Run SQL statements on Snowflake
Advanced Settings
Advanced Settings
- System Prompt: Customize the agent’s behavior and SQL guidelines.
- Model: Select the language model for the agent (must support function calling).
- Max Steps: Set the maximum number of steps the agent can take (1-20).
- Enable Model Fallbacks: Allow automatic fallback to alternative models if the primary model fails.
Expected Inputs and Outputs
-
Inputs:
- The node accepts text input that can be referenced in the prompt using format strings.
-
Outputs:
- output: The final result from the agent’s actions
- conversation: Detailed log of the agent’s reasoning and SQL queries
Use Case Examples
- Automated Reporting: Execute scheduled SQL queries to generate reports and distribute insights to stakeholders.
- Data Analysis: Query large datasets for specific patterns, trends, or anomalies and process the results in workflows.
- Data Integration: Extract data from Snowflake for use in other workflow nodes or export to different systems.
Error Handling and Troubleshooting
- Authentication Issues: Ensure you’ve provided valid Snowflake credentials and your account URL is correct.
- SQL Errors: Verify SQL syntax is correct and tables/columns exist in the specified database and schema.
- Permission Errors: Check that your Snowflake role has the necessary permissions to access the requested data.