Summary of the experience running the Running the Quickstart for dbt and Snowflake Tutorial with dbt Projects on Snowflake
This article summarizes my journey following the official Quickstart for dbt and Snowflake tutorial adapted for dbt Projects on Snowflake. The goal is to learn dbt from scratch and see how its direct integration works within Snowflake without using dbt Cloud.
Important notes on scope
dbt Cloud-exclusive features such as Managed Repository, integrated documentation viewing, and dbt Cloud-specific environment and job configuration are omitted. In dbt Projects on Snowflake, environment control is managed through profiles.yml and deployment is done as a dbt project within a deployable schema that can be run as a task.
Prerequisites
You will need a trial account in Snowflake. Activate secondary roles if necessary by checking the default_secondary_roles property for your user and adjusting it otherwise.
Step 1 Introduction
The tutorial flow covers creating a worksheet in Snowflake, loading sample data, converting queries into dbt models, declaring sources, adding tests, documenting models, and finally deploying the project for scheduled execution via tasks in Snowflake.
Step 2 Create a worksheet
From Projects in Snowflake, a new Worksheet is created to run queries and administrative operations. This will be the space from which databases, schemas, tables are created and tasks are defined.
Step 3 Load data
The tutorial includes creating a raw database with jaffle_shop and stripe schemas and loading public CSVs into sample tables. In this phase, the customers, orders, and payment tables are generated to be used as source data.
Step 4 Connect dbt to Snowflake
In dbt Cloud this requires specific configuration, but in dbt Projects on Snowflake it is not necessary to establish an external connection since development runs within the Snowflake environment.
Step 5 Managed Repository
Omitted. In this tutorial, development is done without Git integration. If Git is desired, it can be integrated with dbt Projects on Snowflake using the appropriate flows and external repositories.
Step 6 Initialize dbt project and start developing
From Projects Workspaces, a new dbt Project is created by specifying name, role, warehouse, database, and development schema. This information is reflected in profiles.yml and allows running dbt commands like dbt run from the Workspace to materialize models in the development schema.
Step 7 Build the first model
A SQL file customers.sql is created in the models folder that combines customers and orders to produce a view or table of customers with basic metrics. Running dbt run materializes the model according to the active configuration.
Step 8 Change materialization
Materialization defines whether a model is created as a view, table, incremental, or ephemeral. It can be adjusted globally in dbt_project.yml or per file using config. Changing from view to table exemplifies how dbt recreates objects in Snowflake according to the materialization policy.
Step 9 Remove sample models
Deleting SQL files from the project does not automatically delete existing tables or views in Snowflake. Objects must be cleaned up manually if you want to free up space or avoid confusion.
Step 10 Models on models
Following best practices, staging models stg_customers and stg_orders are created to normalize and rename columns. The final customers file references those staging models using the ref function to declare dependencies. Compiling the project allows viewing the dependency DAG in the Workspace.
Step 11 Models on sources
Sources are declared in a YAML file to indicate the raw database and source schemas. Then the staging models use the source function to reference the original tables. When compiling, the source line appears in the DAG and references resolve to the physical table name.
Step 12 Add tests
dbt allows defining data tests in YAML files. In the example, unique, not_null, accepted_values, and relationships tests are defined to ensure key integrity and valid values. Running dbt test executes these checks and shows PASS or FAIL in the Workspace logs.
Step 13 Document models
Generating documentation with dbt docs generate was not available in the version of dbt Projects on Snowflake used in my test, so that part was omitted. In environments with support, dbt docs allows browsing models, tests, and lineage with descriptions and examples.
Step 14 Commit changes
In this tutorial we develop without Git. To integrate version control, you need to configure an external repository and link it to the dbt Project following the Git integration guide for dbt Projects on Snowflake.
Step 15 Deploy dbt
For production, a prod target is added in profiles.yml with the PROD output schema. The dbt project is deployed as a dbt Project in a deployable schema, and from there a Snowflake task is created that runs dbt build with target prod. The task can be scheduled and run manually to materialize tables and run tests in production.
Conclusion
Running the Quickstart for dbt and Snowflake within dbt Projects on Snowflake is an excellent way to learn dbt while taking advantage of native execution in Snowflake. The flow covers everything from data loading to automated deployment as a task, with control of materializations, tests, and lineage.
About Q2BSTUDIO
At Q2BSTUDIO, we are a software development company specialized in custom applications and custom software solutions for businesses of all sizes. Our team combines expertise in artificial intelligence and backend and frontend development to deliver scalable and secure products. We offer cybersecurity services to protect applications and data, AWS and Azure cloud services to deploy resilient infrastructures, and business intelligence services that include Power BI integration to create actionable dashboards.
Our artificial intelligence capabilities encompass AI solutions for businesses, design and implementation of AI agents, and custom models that automate processes and extract value from data. We integrate data pipelines with tools like dbt and cloud platforms to deliver reproducible and auditable processes. If you are looking for a partner to create custom applications or transform your data into business intelligence, Q2BSTUDIO brings technical expertise and a results-oriented approach.
Keywords and positioning
Custom applications, custom software, artificial intelligence, cybersecurity, AWS and Azure cloud services, business intelligence services, AI for businesses, AI agents, Power BI, custom applications, custom software, artificial intelligence
Final invitation
If you want to try the Quickstart flow and need advice to integrate it into your data pipelines or deploy solutions with security and scalability, contact Q2BSTUDIO and we will help you design the architecture, implement models, and take your data to production.
