Overview

Note: DCM Projects is currently in Public Preview. See the DCM Projects documentation for the latest details.

Prerequisites

What You'll Learn

What You'll Need

What You'll Build

Create a Workspace from Git

Set Up Roles and Permissions

1. Create a DCM Developer Role

USE ROLE ACCOUNTADMIN;

CREATE ROLE IF NOT EXISTS dcm_developer;
SET user_name = (SELECT CURRENT_USER());
GRANT ROLE dcm_developer TO USER IDENTIFIER($user_name);

2. Grant Infrastructure Privileges

GRANT CREATE WAREHOUSE ON ACCOUNT TO ROLE dcm_developer;
GRANT CREATE ROLE ON ACCOUNT TO ROLE dcm_developer;
GRANT CREATE DATABASE ON ACCOUNT TO ROLE dcm_developer;
GRANT EXECUTE MANAGED TASK ON ACCOUNT TO ROLE dcm_developer;
GRANT EXECUTE TASK ON ACCOUNT TO ROLE dcm_developer;
GRANT MANAGE GRANTS ON ACCOUNT TO ROLE dcm_developer;

3. Grant Data Quality Privileges

GRANT APPLICATION ROLE SNOWFLAKE.DATA_QUALITY_MONITORING_VIEWER TO ROLE dcm_developer;
GRANT APPLICATION ROLE SNOWFLAKE.DATA_QUALITY_MONITORING_ADMIN TO ROLE dcm_developer;
GRANT DATABASE ROLE SNOWFLAKE.DATA_METRIC_USER TO ROLE dcm_developer;
GRANT EXECUTE DATA METRIC FUNCTION ON ACCOUNT TO ROLE dcm_developer;

4. Create the DCM Project Object

USE ROLE dcm_developer;

CREATE DATABASE IF NOT EXISTS dcm_demo;
CREATE SCHEMA IF NOT EXISTS dcm_demo.projects;

CREATE OR REPLACE DCM PROJECT dcm_demo.projects.dcm_project_dev
    COMMENT = 'for testing DCM Projects Quickstarts';

Note: After running this script, refresh your browser so Snowsight picks up the newly created DCM Project object. It won't appear in the Workspaces project selector until you do.

CLI Alternative: You can also create the DCM Project object from the command line using Snowflake CLI:

snow dcm create --target DCM_DEV

Explore the Project Files

Manifest

manifest_version: 2
type: DCM_PROJECT

default_target: DCM_DEV

targets:
  DCM_DEV:
    account_identifier: MYORG-MY_DEV_ACCOUNT        # update to your account identifier
    project_name: DCM_DEMO.PROJECTS.DCM_PROJECT_DEV
    project_owner: DCM_DEVELOPER
    templating_config: DEV

  DCM_STAGE:
    account_identifier: MYORG-MY_STAGE_ACCOUNT
    project_name: DCM_DEMO.PROJECTS.DCM_PROJECT_STG
    project_owner: DCM_STAGE_DEPLOYER
    templating_config: STAGE

  DCM_PROD_US:
    account_identifier: MYORG-MY_ACCOUNT_US
    project_name: DCM_DEMO.PROJECTS.DCM_PROJECT_PROD
    project_owner: DCM_PROD_DEPLOYER
    templating_config: PROD

templating:
  defaults:
    user: "GITHUB_ACTIONS_SERVICE_USER"
    wh_size: "X-SMALL"

  configurations:
    DEV:
      env_suffix: "_DEV"
      user: "INSERT_YOUR_USER"                          # update to your username
      project_owner_role: "DCM_DEVELOPER"
      teams:
        - name: "DEV_TEAM_1"
          data_retention_days: 1
          needs_sandbox_schema: true

    PROD:
      env_suffix: ""
      wh_size: "LARGE"
      project_owner_role: "DCM_PROD_DEPLOYER"
      teams:
        - name: "Marketing"
          data_retention_days: 1
          needs_sandbox_schema: true
        - name: "Finance"
          data_retention_days: 30
          needs_sandbox_schema: false
        - name: "HR"
          data_retention_days: 7
          needs_sandbox_schema: false
        - name: "IT"
          data_retention_days: 14
          needs_sandbox_schema: true
        - name: "Sales"
          data_retention_days: 1
          needs_sandbox_schema: false
        - name: "Research"
          data_retention_days: 7
          needs_sandbox_schema: true

Important: Before running a Plan, update account_identifier (line 8) and user (line 28) under the DCM_DEV target in manifest.yml to match your Snowflake account. The last query in scripts/01_pre_deploy.sql (step 6) returns both values — copy them from that output.

Definition Files

DEFINE DATABASE dcm_demo_1{{env_suffix}}
    COMMENT = 'This is a Quickstart Demo for DCM Projects';

DEFINE SCHEMA dcm_demo_1{{env_suffix}}.raw;

DEFINE TABLE dcm_demo_1{{env_suffix}}.raw.menu (
    menu_item_id NUMBER,
    menu_item_name VARCHAR,
    item_category VARCHAR,
    cost_of_goods_usd NUMBER(10, 2),
    sale_price_usd NUMBER(10, 2)
)
CHANGE_TRACKING = TRUE;

DEFINE DYNAMIC TABLE dcm_demo_1{{env_suffix}}.analytics.enriched_order_details
WAREHOUSE = dcm_demo_1_wh{{env_suffix}}
TARGET_LAG = 'DOWNSTREAM'
INITIALIZE = 'ON_SCHEDULE'
DATA_METRIC_SCHEDULE = 'TRIGGER_ON_CHANGES'
AS
SELECT
    oh.order_id,
    oh.order_ts,
    od.quantity,
    m.menu_item_name,
    m.item_category,
    m.sale_price_usd,
    (od.quantity * m.sale_price_usd) AS line_item_revenue,
    (od.quantity * (m.sale_price_usd - m.cost_of_goods_usd)) AS line_item_profit,
    c.customer_id,
    c.first_name,
    c.last_name,
    INITCAP(c.city) AS customer_city,
    t.truck_id,
    t.truck_brand_name
FROM dcm_demo_1{{env_suffix}}.raw.order_header oh
JOIN dcm_demo_1{{env_suffix}}.raw.order_detail od ON oh.order_id = od.order_id
JOIN dcm_demo_1{{env_suffix}}.raw.menu m ON od.menu_item_id = m.menu_item_id
JOIN dcm_demo_1{{env_suffix}}.raw.customer c ON oh.customer_id = c.customer_id
JOIN dcm_demo_1{{env_suffix}}.raw.truck t ON oh.truck_id = t.truck_id;

Why INITIALIZE = ON_SCHEDULE? By default, dynamic tables use INITIALIZE = ON_CREATE, which runs a full synchronous refresh during the CREATE statement. When a project creates many tables at once, those synchronous refreshes can make deployments slow. INITIALIZE = ON_SCHEDULE skips the refresh at creation time and lets DCM finish the deployment quickly — each dynamic table populates itself in the background on its normal schedule. This is the recommended pattern for DCM deployments where speed and predictability matter most.

Macros

{% macro create_team_roles(team) %}

    DEFINE ROLE {{team}}_OWNER{{env_suffix}};
    DEFINE ROLE {{team}}_DEVELOPER{{env_suffix}};
    DEFINE ROLE {{team}}_USAGE{{env_suffix}};

    GRANT USAGE ON DATABASE dcm_demo_1{{env_suffix}}
        TO ROLE {{team}}_USAGE{{env_suffix}};
    GRANT OWNERSHIP ON SCHEMA dcm_demo_1{{env_suffix}}.{{team}}
        TO ROLE {{team}}_OWNER{{env_suffix}};

    GRANT CREATE DYNAMIC TABLE, CREATE TABLE, CREATE VIEW
        ON SCHEMA dcm_demo_1{{env_suffix}}.{{team}}
        TO ROLE {{team}}_DEVELOPER{{env_suffix}};

    GRANT ROLE {{team}}_USAGE{{env_suffix}} TO ROLE {{team}}_DEVELOPER{{env_suffix}};
    GRANT ROLE {{team}}_DEVELOPER{{env_suffix}} TO ROLE {{team}}_OWNER{{env_suffix}};
    GRANT ROLE {{team}}_OWNER{{env_suffix}} TO ROLE {{project_owner_role}};

{% endmacro %}

{% for team in teams %}
    {% set team_name = team.name | upper %}

    DEFINE SCHEMA dcm_demo_1{{env_suffix}}.{{team_name}}
        COMMENT = 'using JINJA dictionary values'
        DATA_RETENTION_TIME_IN_DAYS = {{ team.data_retention_days }};

    {{ create_team_roles(team_name) }}

    {% if team.needs_sandbox_schema | default(false) %}
        DEFINE SCHEMA dcm_demo_1{{env_suffix}}.{{team_name}}_SANDBOX
            COMMENT = 'Sandbox schema defined via dictionary flag'
            DATA_RETENTION_TIME_IN_DAYS = 1;
    {% endif %}

{% endfor %}

Plan and Deploy the Pipeline

Select the Project

CLI Alternative: From the command line, plan with:

snow dcm plan --target DCM_DEV -D "user='YOUR_USERNAME'"

Execute the Plan

Review the Plan Output

Deploy

CLI Alternative: From the command line, deploy with:

snow dcm deploy --target DCM_DEV --alias "Initial pipeline deployment"

Insert Sample Data

1. Insert Sample Data

2. Refresh Dynamic Tables

EXECUTE DCM PROJECT DCM_DEMO.PROJECTS.DCM_PROJECT_DEV REFRESH ALL;

3. Verify

SELECT * FROM dcm_demo_1_dev.serve.v_dashboard_daily_sales;

SELECT * FROM dcm_demo_1_dev.analytics.enriched_order_details;
SELECT * FROM dcm_demo_1_dev.analytics.menu_item_popularity;
SELECT * FROM dcm_demo_1_dev.analytics.customer_spending_summary;

Cleanup

USE ROLE dcm_developer;
EXECUTE DCM PROJECT dcm_demo.projects.dcm_project_dev PURGE;

DROP DCM PROJECT IF EXISTS dcm_demo.projects.dcm_project_dev;
DROP SCHEMA IF EXISTS dcm_demo.projects;
DROP DATABASE IF EXISTS dcm_demo;

USE ROLE ACCOUNTADMIN;
DROP ROLE IF EXISTS dcm_developer;

Conclusion and Resources

What's Next

Related Resources