At the beginning of this year, our team was afforded a unique opportunity: to rebuild our existing dbt project from scratch.

Launched in mid-2019, the legacy project had grown organically into a tangled mess of 1800+ models, with further development becoming more and more difficult.

Faced with the challenge of retroactively imposing order on the existing project, our leadership gave us the opportunity to start fresh instead.

We jumped at the chance, and began applying many of the lessons we learned at Coalesce 2022 to set the new project up for success:
- SQL linting, with SQLFluff
- YAML linting, with yamllint
- dbt best practices, with dbt-checkpoint and dbt-project-evaluator

As a result, this core project has become the model for multi-project collaboration at Cityblock. Rather than a single monolithic project, the new state features a collection of smaller projects, each governed by a high bar for code quality. 

<span data-sheets-formula="='San Diego - approved'!R[0]C[6]" data-sheets-userformat='{"2":256,"11":4}' data-sheets-value='{"1":2,"2":"At the beginning of this year, our team was afforded a unique opportunity: to rebuild our existing dbt project from scratch.\n\nLaunched in mid-2019, the legacy project had grown organically into a tangled mess of 1800+ models, with further development becoming more and more difficult.\n\nFaced with the challenge of retroactively imposing order on the existing project, our leadership gave us the opportunity to start fresh instead.\n\nWe jumped at the chance, and began applying many of the lessons we learned at Coalesce 2022 to set the new project up for success:\n- SQL linting, with SQLFluff\n- YAML linting, with yamllint\n- dbt best practices, with dbt-checkpoint and dbt-project-evaluator\n\nAs a result, this core project has become the model for multi-project collaboration at Cityblock. Rather than a single monolithic project, the new state features a collection of smaller projects, each governed by a high bar for code quality. "}'>At the beginning of this year, our team was afforded a unique opportunity: to rebuild our existing dbt project from scratch. Launched in mid-2019, the legacy project had grown organically into a tangled mess of 1800+ models, with further development becoming more and more difficult. Faced with the challenge of retroactively imposing order on the existing project, our leadership gave us the opportunity to start fresh instead. We jumped at the chance, and began applying many of the lessons we learned at Coalesce 2022 to set the new project up for success: - SQL linting, with SQLFluff - YAML linting, with yamllint - dbt best practices, with dbt-checkpoint and dbt-project-evaluator As a result, this core project has become the model for multi-project collaboration at Cityblock. Rather than a single monolithic project, the new state features a collection of smaller projects, each governed by a high bar for code quality.

San Diego

Online

Staff analytics engineer

Analytics Engineer

One to many: Moving from a monolithic dbt project to multi-project collaboration

Community

Live streamed

At the beginning of this year, our team was afforded a unique opportunity: to rebuild our existing dbt project from scratch.

Launched in mid-2019, the legacy project had grown organically into a...

Katie Claiborne

About

Schedule

Build your path at Coalesce