Deploy

Deploy your PyRel model to make its outputs available to others in Snowflake. Deployment exposes output views that applications and colleagues can use.

Where to start

• Configure deployment

  Choose the Snowflake schema, role, and refresh settings for your model.

  • Set the deployment target and role.
  • Choose how and when model outputs refresh.

  Configure deployment

• Deploy a model

  Deploy a configured PyRel model and make its outputs available in Snowflake.

  • Check that your model and deployment settings are ready.
  • Run the deployment and verify the outputs.

  Deploy a model

How deployment works

Your PyRel project

• Semantic model
• Deployment settings

compiles into

Your Snowflake account

Deployment schema

Model outputs

• Views
• Tables
• Dynamic tables

Refresh and management

• Procedures
• Tasks

compute outputs

trigger refresh

RelationalAI Native App

Reasoners

• Logic
• Predictive
• Prescriptive

access through

• SQL
• Python
• Rel App
• MCP server

A deployed model turns your Python definitions into shared Snowflake objects. Refreshes keep your model's outputs up to date with source data by running its computations on reasoners in the RelationalAI Native App. Query its outputs with SQL or use the model through Python, the Rel app, and connected agents.

Explore the documentation

Configure deploymentDeploy a modelMonitor a deployed modelCollaborate on a deployed modelTroubleshoot a deployed modelOperate RelationalAIManage reasoners